Predicting Customer Purchase Value and Product Price in E-Commerce Transactions Using R
WQD7004 Programming for Data Science — Group Project
14 June 2026
Project Overview
Dataset: E-commerce Purchase History from Electronics Store (Source: Kaggle)
Project Objectives
- To explore and clean a large-scale e-commerce purchase history dataset.
- To analyse customer purchase behaviour based on product, brand, category, and price.
- To build a classification model to predict whether a purchase is high-value or low-value.
- To build a regression model to predict product price.
1.0 Data Exploration
1.1 Install Required Packages
The packages below only need to be installed once. This chunk is set
to eval = FALSE so it does not re-run every time the
document is knitted.
1.2 Load Libraries
1.3 Load Dataset (Local File)
The dataset is loaded directly from the local working directory.
Place kz.csv (raw data) or kz_cleaned.csv
(cleaned data) in the same folder as this .Rmd file before
knitting. The chunk reads the raw file if available, otherwise it falls
back to the cleaned file. The cleaning steps in Section 2.0 are safe to
run on either version.
col_spec <- cols(
event_time = col_character(),
order_id = col_character(),
product_id = col_character(),
category_id = col_character(),
user_id = col_character()
)
if (file.exists("kz.csv")) {
df <- read_csv("kz.csv", col_types = col_spec)
message("Loaded raw dataset: kz.csv")
} else {
df <- read_csv("kz_cleaned.csv", col_types = col_spec)
message("Loaded cleaned dataset: kz_cleaned.csv")
}
glimpse(df)## Rows: 2,633,521
## Columns: 8
## $ event_time <chr> "2020-04-24 11:50:39 UTC", "2020-04-24 11:50:39 UTC", "2…
## $ order_id <chr> "2294359932054536986", "2294359932054536986", "229444402…
## $ product_id <chr> "1515966223509089906", "1515966223509089906", "227394831…
## $ category_id <chr> "2268105426648170900", "2268105426648170900", "226810543…
## $ category_code <chr> "electronics.tablet", "electronics.tablet", "electronics…
## $ brand <chr> "samsung", "samsung", "huawei", "huawei", "karcher", "ma…
## $ price <dbl> 162.01, 162.01, 77.52, 77.52, 217.57, 39.33, 1387.01, 13…
## $ user_id <chr> "1515915625441993984", "1515915625441993984", "151591562…1.4 Initial Inspection
## [1] 2633521 8## spc_tbl_ [2,633,521 × 8] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
## $ event_time : chr [1:2633521] "2020-04-24 11:50:39 UTC" "2020-04-24 11:50:39 UTC" "2020-04-24 14:37:43 UTC" "2020-04-24 14:37:43 UTC" ...
## $ order_id : chr [1:2633521] "2294359932054536986" "2294359932054536986" "2294444024058086220" "2294444024058086220" ...
## $ product_id : chr [1:2633521] "1515966223509089906" "1515966223509089906" "2273948319057183658" "2273948319057183658" ...
## $ category_id : chr [1:2633521] "2268105426648170900" "2268105426648170900" "2268105430162997728" "2268105430162997728" ...
## $ category_code: chr [1:2633521] "electronics.tablet" "electronics.tablet" "electronics.audio.headphone" "electronics.audio.headphone" ...
## $ brand : chr [1:2633521] "samsung" "samsung" "huawei" "huawei" ...
## $ price : num [1:2633521] 162 162 77.5 77.5 217.6 ...
## $ user_id : chr [1:2633521] "1515915625441993984" "1515915625441993984" "1515915625447879434" "1515915625447879434" ...
## - attr(*, "spec")=
## .. cols(
## .. event_time = col_character(),
## .. order_id = col_character(),
## .. product_id = col_character(),
## .. category_id = col_character(),
## .. category_code = col_character(),
## .. brand = col_character(),
## .. price = col_double(),
## .. user_id = col_character()
## .. )
## - attr(*, "problems")=<externalptr>## [1] "event_time" "order_id" "product_id" "category_id"
## [5] "category_code" "brand" "price" "user_id"1.5 Check Missing Values
## event_time order_id product_id category_id category_code
## 0 0 0 431954 612202
## brand price user_id
## 506005 431954 2069352## event_time order_id product_id category_id category_code
## 0.00000 0.00000 0.00000 16.40215 23.24652
## brand price user_id
## 19.21401 16.40215 78.577391.6 Check Duplicates
## [1] 675Interpretation
The initial exploration reveals two data quality issues that must be
addressed before analysis: missing values concentrated in identifier and
descriptive columns (category_code, brand,
user_id), and a number of fully duplicated rows. Both
issues are handled in the cleaning stage below.
2.0 Data Cleaning
The cleaning process performs four operations: removing duplicate
rows, converting event_time from a text string to a proper
datetime object, converting category_id to character, and
replacing missing values with sensible placeholders (“Unknown” for
categorical fields, 0 for price).
df_cleaned <- df %>%
# 1. Remove duplicate rows
distinct() %>%
# 2. Fix the event_time format to proper datetime
mutate(event_time = ymd_hms(event_time)) %>%
# 3. Change category_id to character
mutate(category_id = as.character(category_id)) %>%
# 4. Handle missing values accurately based on column types
mutate(
category_id = ifelse(is.na(category_id), "Unknown", category_id),
category_code = ifelse(is.na(category_code), "Unknown", category_code),
brand = ifelse(is.na(brand), "Unknown", brand),
user_id = ifelse(is.na(user_id), "Unknown", user_id),
price = ifelse(is.na(price), 0, price)
)
# FINAL VERIFICATION CHECK
print("--- FINAL VERIFICATION CHECKS ---")## [1] "--- FINAL VERIFICATION CHECKS ---"## [1] "Remaining Duplicates: 0"## [1] "Remaining NAs per column (Target is ALL ZEROS):"## event_time order_id product_id category_id category_code
## 0 0 0 0 0
## brand price user_id
## 0 0 0Interpretation
After cleaning, the dataset contains no duplicates and no missing
values. Categorical gaps were filled with “Unknown” rather than dropped,
preserving the full transaction volume for analysis. The
event_time column is now a proper datetime, which enables
the time-based feature engineering used later in the classification
model.
3.0 Exploratory Data Analysis
3.1 Statistical Summary
Statistical summary was generated to understand the distribution and characteristics of each variable before further analysis.
| Name | df_cleaned |
| Number of rows | 2632846 |
| Number of columns | 8 |
| _______________________ | |
| Column type frequency: | |
| character | 6 |
| numeric | 1 |
| POSIXct | 1 |
| ________________________ | |
| Group variables | None |
Variable type: character
| skim_variable | n_missing | complete_rate | min | max | empty | n_unique | whitespace |
|---|---|---|---|---|---|---|---|
| order_id | 0 | 1 | 19 | 19 | 0 | 1435266 | 0 |
| product_id | 0 | 1 | 19 | 19 | 0 | 25113 | 0 |
| category_id | 0 | 1 | 7 | 19 | 0 | 928 | 0 |
| category_code | 0 | 1 | 4 | 38 | 0 | 511 | 0 |
| brand | 0 | 1 | 2 | 19 | 0 | 23022 | 0 |
| user_id | 0 | 1 | 7 | 19 | 0 | 233835 | 0 |
Variable type: numeric
| skim_variable | n_missing | complete_rate | mean | sd | p0 | p25 | p50 | p75 | p100 | hist |
|---|---|---|---|---|---|---|---|---|---|---|
| price | 0 | 1 | 128.81 | 228.44 | 0 | 4.61 | 34.7 | 161.78 | 50925.9 | ▇▁▁▁▁ |
Variable type: POSIXct
| skim_variable | n_missing | complete_rate | min | max | median | n_unique |
|---|---|---|---|---|---|---|
| event_time | 0 | 1 | 1970-01-01 00:33:40 | 2020-11-21 10:10:30 | 2020-06-08 08:23:13 | 1316174 |
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.00 4.61 34.70 128.81 161.78 50925.903.2 Price Distribution
The distribution of product prices was examined to understand the spread of transaction values across the store.
# Drop non-positive (spending distribution of real transactions)
price_positive <- df_cleaned %>%
filter(price > 0)
# Raw price scale
ggplot(price_positive, aes(x = price)) +
geom_histogram(bins = 50, fill = "steelblue", color = "white") +
labs(
title = "Distribution of Product Prices",
x = "Price",
y = "Number of Transactions"
)
# Log10 scale
ggplot(price_positive, aes(x = price)) +
geom_histogram(bins = 50, fill = "steelblue", color = "white") +
scale_x_log10() +
labs(
title = "Distribution of Product Prices (Log10 Scale)",
x = "Price (log10)",
y = "Number of Transactions"
)
Interpretation
On the raw scale almost all transactions cluster near the low end with a long tail of expensive products. On the log scale the distribution is roughly bell-shaped, confirming that most purchases fall within a moderate price band while high-value purchases are comparatively rare. This pattern directly supports the high-value vs low-value classification target model.
3.3 Top Brands by Transaction Volume
The most frequently purchased brands were identified to understand which brands drive transaction volume.
# Count transactions per brand - 10 most frequent.
# Exclude "Unknown" brands
brand_freq <- df_cleaned %>%
filter(brand != "Unknown") %>%
count(brand, sort = TRUE) %>%
head(10)
# reorder() sorts the bars by count; coord_flip() to keep the labels readable
ggplot(brand_freq, aes(x = reorder(brand, n), y = n)) +
geom_col(fill = "darkorange") +
coord_flip() +
labs(
title = "Top 10 Brands by Transaction Volume",
x = "Brand",
y = "Number of Transactions"
)
Interpretation
A small group of brands accounts for a large share of all transactions. This concentration is why the modelling sections collapse the long tail of rare brands into an “Other” group and keep only the most frequent brands as predictors.
3.4 Top Product Categories by Transaction Volume
The most common product categories were examined to see where purchasing activity is concentrated.
# Count frequent product categories - 10 most frequent.
# Exclude "Unknown" categories
category_freq <- df_cleaned %>%
filter(category_code != "Unknown") %>%
count(category_code, sort = TRUE) %>%
head(10)
ggplot(category_freq, aes(x = reorder(category_code, n), y = n)) +
geom_col(fill = "darkorange") +
coord_flip() +
labs(
title = "Top 10 Product Categories by Transaction Volume",
x = "Category",
y = "Number of Transactions"
)
Interpretation
Purchasing activity is dominated by a handful of categories (largely electronics and accessories). The remaining categories each contribute only a small fraction of transactions. Hence, rare categories require grouping into “Other” before modelling.
3.5 Average Price by Category
Average prices were compared across the categories to see how product value differs by category.
# Average prices of known categories
avg_price_category <- df_cleaned %>%
filter(category_code != "Unknown", price > 0) %>%
group_by(category_code) %>%
summarise(
avg_price = mean(price),
n_transactions = n()
) %>%
arrange(desc(n_transactions)) %>%
head(10)
# Sort highest-volume categories
ggplot(avg_price_category, aes(x = reorder(category_code, avg_price), y = avg_price)) +
geom_col(fill = "seagreen") +
coord_flip() +
labs(
title = "Average Price Among Top 10 Categories",
x = "Category",
y = "Average Price"
)
Interpretation
Average price varies widely between categories: high-end electronics such as smartphones and notebooks sit far above accessories and consumables. This shows that category is a meaningful signal for price, suitable to be used as a predictor in both the classification and regression models.
3.6 Price Spread Across Top Categories
A boxplot was used to compare not just the average price but the full spread of prices within each major category.
# Top categories from the plot above
top_category_names <- avg_price_category$category_code
price_by_category <- df_cleaned %>%
filter(category_code %in% top_category_names, price > 0)
# Log scale
ggplot(price_by_category, aes(x = category_code, y = price)) +
geom_boxplot(fill = "lightblue") +
scale_y_log10() +
theme(axis.text.x = element_text(angle = 90, hjust = 1)) +
labs(
title = "Price Distribution Across Top Categories (Log10 Scale)",
x = "Category",
y = "Price (log10)"
)
Interpretation
The price range varies across product categories. Some categories have similar prices, while others contain both low-priced and expensive products with many high outliers. This shows that category alone is not enough to accurately predict product price. Therefore, additional variables such as brand need to be included in the regression model to improve prediction accuracy.
3.7 Purchasing Patterns Over Time
Transaction timing was explored to reveal when customers are most active. These time-based patterns are used to engineer the temporal features.
# Drop any rows that failed to parse to avoid blank categories in the plots
time_features <- df_cleaned %>%
filter(!is.na(event_time)) %>%
mutate(
hour = hour(event_time),
weekday = wday(event_time, label = TRUE, week_start = 1),
month = month(event_time, label = TRUE)
)
# Transactions by hour of day
ggplot(time_features, aes(x = hour)) +
geom_bar(fill = "purple") +
labs(
title = "Transactions by Hour of Day",
x = "Hour (0-23)",
y = "Number of Transactions"
)
# Transactions by weekday
ggplot(time_features, aes(x = weekday)) +
geom_bar(fill = "purple") +
labs(
title = "Transactions by Weekday",
x = "Weekday",
y = "Number of Transactions"
)
# Transactions by month
ggplot(time_features, aes(x = month)) +
geom_bar(fill = "purple") +
labs(
title = "Transactions by Month",
x = "Month",
y = "Number of Transactions"
)
Interpretation
Purchases are not spread evenly across time: activity rises and falls
across the hours of the day and differs between weekdays and months.
Because these patterns are clear, hour,
weekday and month can be engineered as
features for the classification model.
3.8 EDA Summary
The exploratory analysis produced several findings that shape the modelling work that follows:
- Price is strongly right-skewed, with most transactions at low values and a small number of high-value purchases. This supports the high-value vs low-value target created in Section 4.0.
- Brands and categories are highly concentrated, where a few brands and categories account for most transactions. This justifies grouping rare levels into an “Other” category before modelling.
- Category and brand are clearly linked to price, both in average level and in spread, making them useful predictors for the regression model in Section 5.0.
- Purchasing activity varies by hour, weekday and month, which motivates the time-based features used in the classification model.
4.0 Classification Modelling
4.1 Take Sample Data
A sample is drawn to reduce computation time. The original
df is preserved; the sample is stored in
df_class for the classification work.
Interpretation
The dataset was reduced to 50,000 rows, which ensures faster computation and efficiency while still keeping a large enough sample to be representative.
4.2 Feature Engineering
df_class$hour <- as.factor(format(df_class$event_time, "%H"))
df_class$weekday <- as.factor(weekdays(df_class$event_time))
df_class$month <- as.factor(format(df_class$event_time, "%m"))Interpretation
Time-based features may reveal purchase behaviour patterns, such as high-value purchases being more common at certain hours or days.
4.3 Create Target Variable
## 75%
## 157.15df_class$purchase_class <- ifelse(
df_class$price >= threshold,
"High",
"Low"
)
df_class$purchase_class <- as.factor(df_class$purchase_class)
table(df_class$purchase_class)##
## High Low
## 12503 37497Interpretation
To accurately predict customer behaviour and isolate high-value
revenue drivers, a binary target classification variable
(purchase_class) was engineered from the continuous
price metric. Following an exploratory baseline audit of
the price distribution, a strict mathematical threshold was established
at the 75th percentile (3rd quartile).
Any transaction containing an item priced at or above the threshold was categorised as “High” (representing the premium top-tier quadrant of sales), while items priced under the threshold were flagged as “Low”.
The resulting distribution shows roughly 12,500 high-value vs 37,500 low-value purchases, indicating a moderately imbalanced dataset suitable for classification modelling.
4.4 Keep Top 20 Brands/Categories
# Top brands
top_brands <- names(
sort(table(df_class$brand), decreasing = TRUE)
)[1:20]
df_class$brand <- ifelse(
df_class$brand %in% top_brands,
df_class$brand,
"Other"
)
# Top categories
top_categories <- names(
sort(table(df_class$category_code), decreasing = TRUE)
)[1:20]
df_class$category_code <- ifelse(
df_class$category_code %in% top_categories,
df_class$category_code,
"Other"
)
# Convert to factor
df_class$brand <- as.factor(df_class$brand)
df_class$category_code <- as.factor(df_class$category_code)Interpretation
This reduces noise and avoids sparse categories, making the model more efficient and focused on the most common brands/categories.
4.5 Train/Test Split
# Select variables
model_df <- df_class %>%
select(
purchase_class,
brand,
category_code,
hour,
weekday,
month
)
# Train/Test Split (80:20)
set.seed(123)
trainIndex <- createDataPartition(
model_df$purchase_class,
p = 0.80,
list = FALSE
)
data_train <- model_df[trainIndex, ]
data_test <- model_df[-trainIndex, ]4.6 Cross Validation
# Repeated CV, 5 folds
ctrl <- trainControl(
method = "repeatedcv",
number = 5,
classProbs = TRUE,
summaryFunction = twoClassSummary
)Interpretation
- Repeated cross-validation with class probabilities and ROC metrics was set up.
- This provides a more reliable estimate of model performance by testing across multiple folds, and using ROC/AUC ensures evaluation remains sound even with imbalanced classes.
4.7 Logistic Regression Model
model_log <- train(
purchase_class ~ .,
data = data_train,
method = "glm",
family = "binomial",
metric = "ROC",
trControl = ctrl
)
# Display model
print(model_log)## Generalized Linear Model
##
## 40001 samples
## 5 predictor
## 2 classes: 'High', 'Low'
##
## No pre-processing
## Resampling: Cross-Validated (5 fold, repeated 1 times)
## Summary of sample sizes: 32002, 32001, 32000, 32001, 32000
## Resampling results:
##
## ROC Sens Spec
## 0.9123125 0.7338785 0.9108608Interpretation
- ROC ≈ 0.912 → The model has excellent ability to distinguish between High vs Low value purchases.
- Sensitivity ≈ 0.734 → About 73% of high-value purchases were correctly identified.
- Specificity ≈ 0.911 → About 91% of low-value purchases were correctly identified.
- The model is strong overall, but slightly better at identifying low-value purchases than high-value ones.
4.8 Naive Bayes Model
model_nb <- train(
purchase_class ~ .,
data = data_train,
method = "nb",
metric = "ROC",
trControl = ctrl
)
# Display model
print(model_nb)## Naive Bayes
##
## 40001 samples
## 5 predictor
## 2 classes: 'High', 'Low'
##
## No pre-processing
## Resampling: Cross-Validated (5 fold, repeated 1 times)
## Summary of sample sizes: 32000, 32000, 32001, 32001, 32002
## Resampling results across tuning parameters:
##
## usekernel ROC Sens Spec
## FALSE NaN NaN NaN
## TRUE 0.8930407 0 1
##
## Tuning parameter 'fL' was held constant at a value of 0
## Tuning
## parameter 'adjust' was held constant at a value of 1
## ROC was used to select the optimal model using the largest value.
## The final values used for the model were fL = 0, usekernel = TRUE and adjust
## = 1.Interpretation
- ROC ≈ 0.893 → Slightly lower than logistic regression (0.912), but still strong discrimination between High vs Low purchases.
- Sensitivity ≈ 0 → The model failed to correctly identify any high-value purchases.
- Specificity ≈ 1 → The model perfectly identified low-value purchases.
- The model is biased toward predicting “Low” purchases only, which makes it unsuitable for detecting high-value customers.
4.9 Random Forest Model
model_rf <- train(
purchase_class ~ .,
data = data_train,
method = "rf",
metric = "ROC",
trControl = ctrl,
ntree = 50
)
# Display model
print(model_rf)## Random Forest
##
## 40001 samples
## 5 predictor
## 2 classes: 'High', 'Low'
##
## No pre-processing
## Resampling: Cross-Validated (5 fold, repeated 1 times)
## Summary of sample sizes: 32000, 32000, 32001, 32002, 32001
## Resampling results across tuning parameters:
##
## mtry ROC Sens Spec
## 2 0.8639198 0.4686595 0.9474966
## 40 0.8937572 0.6827948 0.9171948
## 79 0.8888142 0.6803948 0.9103943
##
## ROC was used to select the optimal model using the largest value.
## The final value used for the model was mtry = 40.Interpretation
- ROC ≈ 0.893 (best at mtry = 40) → Very good discrimination between High vs Low purchases, slightly lower than Logistic Regression (0.912) but stronger than Naive Bayes (0.893 with poor sensitivity).
- Sensitivity ≈ 0.683 → About 68% of high-value purchases correctly identified.
- Specificity ≈ 0.917 → About 92% of low-value purchases correctly identified.
- Random Forest balances both classes better than Naive Bayes, and handles categorical predictors more gracefully than Logistic Regression.
4.10 Model Predictions
4.11 Confusion Matrices
Logistic Regression
## Confusion Matrix and Statistics
##
## Reference
## Prediction High Low
## High 1880 678
## Low 620 6821
##
## Accuracy : 0.8702
## 95% CI : (0.8634, 0.8767)
## No Information Rate : 0.75
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.6565
##
## Mcnemar's Test P-Value : 0.1136
##
## Sensitivity : 0.7520
## Specificity : 0.9096
## Pos Pred Value : 0.7349
## Neg Pred Value : 0.9167
## Prevalence : 0.2500
## Detection Rate : 0.1880
## Detection Prevalence : 0.2558
## Balanced Accuracy : 0.8308
##
## 'Positive' Class : High
## Naive Bayes
## Confusion Matrix and Statistics
##
## Reference
## Prediction High Low
## High 0 0
## Low 2500 7499
##
## Accuracy : 0.75
## 95% CI : (0.7414, 0.7584)
## No Information Rate : 0.75
## P-Value [Acc > NIR] : 0.5054
##
## Kappa : 0
##
## Mcnemar's Test P-Value : <2e-16
##
## Sensitivity : 0.00
## Specificity : 1.00
## Pos Pred Value : NaN
## Neg Pred Value : 0.75
## Prevalence : 0.25
## Detection Rate : 0.00
## Detection Prevalence : 0.00
## Balanced Accuracy : 0.50
##
## 'Positive' Class : High
## Random Forest
## Confusion Matrix and Statistics
##
## Reference
## Prediction High Low
## High 1654 536
## Low 846 6963
##
## Accuracy : 0.8618
## 95% CI : (0.8549, 0.8685)
## No Information Rate : 0.75
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.6156
##
## Mcnemar's Test P-Value : < 2.2e-16
##
## Sensitivity : 0.6616
## Specificity : 0.9285
## Pos Pred Value : 0.7553
## Neg Pred Value : 0.8917
## Prevalence : 0.2500
## Detection Rate : 0.1654
## Detection Prevalence : 0.2190
## Balanced Accuracy : 0.7951
##
## 'Positive' Class : High
## Interpretation
- Logistic Regression is balanced and reliable, slightly favouring low-value detection but still strong for high-value. Naive Bayes collapsed: it only predicts “Low” and is unusable for identifying high-value customers. Random Forest is dependable, slightly less sensitive than Logistic Regression but very strong in specificity.
- Logistic Regression is the best overall model here, with Random Forest as a strong alternative. Naive Bayes is not suitable for this dataset.
4.12 ROC Curve and AUC
# Compare Logistic Regression vs Random Forest
# Logistic Regression probabilities
log_probs <- predict(model_log, data_test, type = "prob")
# Random Forest probabilities
rf_probs <- predict(model_rf, data_test, type = "prob")
# ROC for Logistic Regression
roc_log <- roc(
response = data_test$purchase_class,
predictor = log_probs$High
)
# ROC for Random Forest
roc_rf <- roc(
response = data_test$purchase_class,
predictor = rf_probs$High
)
# Plot Logistic Regression ROC
plot(roc_log, col = "blue", main = "ROC Curve Comparison")
# Add Random Forest ROC
plot(roc_rf, col = "red", add = TRUE)
# Add legend
legend(
"bottomright",
legend = c("Logistic Regression", "Random Forest"),
col = c("blue", "red"),
lwd = 2
)
# AUC values
auc_log <- auc(roc_log)
auc_rf <- auc(roc_rf)
data.frame(
Model = c("Logistic Regression", "Random Forest"),
AUC = round(c(auc_log, auc_rf), 3)
)Interpretation
- Both models perform well above random guessing (AUC > 0.5).
- Logistic Regression edges out Random Forest in overall discrimination (higher AUC).
- Random Forest is more flexible with categorical features and still competitive.
- The ROC curves lying above the diagonal line confirm both models are useful classifiers.
4.13 Feature Importance
## rf variable importance
##
## only 20 most important variables shown (out of 79)
##
## Overall
## category_codeelectronics.smartphone 100.00
## category_codeappliances.kitchen.refrigerators 74.57
## category_codeelectronics.video.tv 52.44
## brandapple 50.12
## category_codecomputers.notebook 45.55
## category_codeappliances.kitchen.washer 44.86
## brandlg 41.29
## brandUnknown 35.14
## brandbeko 33.46
## brandsamsung 30.40
## category_codeUnknown 30.18
## brandOther 29.92
## brandbosch 21.22
## category_codeappliances.kitchen.hood 16.81
## weekdayTuesday 15.28
## weekdaySunday 14.48
## weekdayWednesday 14.35
## weekdaySaturday 14.23
## weekdayThursday 14.15
## category_codeOther 13.49
Interpretation
category_codedominates → Product type is the strongest signal of high vs low purchase value.- Brand matters → Premium brands (Apple, LG, Samsung) are more likely associated with high-value purchases.
- Temporal features (weekday, month) have smaller but non-negligible influence, suggesting purchase behaviour varies by time.
Marketing strategies should prioritise product categories and premium brands, while also considering timing patterns.
4.14 Model Comparison
results <- resamples(
list(
Logistic_Regression = model_log,
Naive_Bayes = model_nb,
Random_Forest = model_rf
)
)
# Summary statistics
summary(results)##
## Call:
## summary.resamples(object = results)
##
## Models: Logistic_Regression, Naive_Bayes, Random_Forest
## Number of resamples: 5
##
## ROC
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## Logistic_Regression 0.9105447 0.9111047 0.9129353 0.9123125 0.9129665 0.9140114
## Naive_Bayes 0.8885272 0.8918082 0.8937953 0.8930407 0.8938679 0.8972051
## Random_Forest 0.8872343 0.8877338 0.8940993 0.8937572 0.8967984 0.9029202
## NA's
## Logistic_Regression 0
## Naive_Bayes 0
## Random_Forest 0
##
## Sens
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## Logistic_Regression 0.7245000 0.7300 0.7321339 0.7338785 0.7321339 0.7506247
## Naive_Bayes 0.0000000 0.0000 0.0000000 0.0000000 0.0000000 0.0000000
## Random_Forest 0.6676662 0.6775 0.6801599 0.6827948 0.6845000 0.7041479
## NA's
## Logistic_Regression 0
## Naive_Bayes 0
## Random_Forest 0
##
## Spec
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## Logistic_Regression 0.9065000 0.9071667 0.9111519 0.9108608 0.9131522 0.9163333
## Naive_Bayes 1.0000000 1.0000000 1.0000000 1.0000000 1.0000000 1.0000000
## Random_Forest 0.9116667 0.9131667 0.9171667 0.9171948 0.9196533 0.9243207
## NA's
## Logistic_Regression 0
## Naive_Bayes 0
## Random_Forest 0
Interpretation
- ROC → Logistic > RF > NB
- Sensitivity → Logistic > RF > NB
- Specificity → RF > Logistic > NB
4.15 Classification Modelling Summary
Logistic Regression is the best overall model, Random Forest is a dependable backup, and Naive Bayes is discarded. Insights show that product type and brand are the key factors influencing high-value purchases.
5.0 Regression Modelling
5.1 Preview Cleaned Dataset
The regression section reuses the cleaned dataset from Section 2.0.
5.2 Data Preparation
5.2.1 Sampling the Dataset
A sample of 150,000 observations was selected to improve computational efficiency during regression modelling.
## [1] 150000 85.2.2 Filtering Valid Brand Values
Invalid or non-meaningful brand values were removed to improve model quality.
## [1] 147838 8Interpretation
The dataset was filtered to retain only records with valid alphabetic brand values. A small number of records with invalid or non-meaningful brand entries were removed to improve the quality of the regression analysis.
5.2.3 Factor Conversion
Categorical variables are converted into factor format for regression modelling.
df_sample$brand <- factor(df_sample$brand)
df_sample$category_code <- factor(df_sample$category_code)
str(df_sample[, c("brand", "category_code")])## tibble [147,838 × 2] (S3: tbl_df/tbl/data.frame)
## $ brand : Factor w/ 605 levels "a-case","accesstyle",..: 251 560 500 82 560 29 91 529 500 137 ...
## $ category_code: Factor w/ 297 levels "0.00","0.05",..: 267 99 183 209 99 267 202 192 183 236 ...Interpretation
Categorical variables were converted to factor format so that they can be correctly handled by the regression model when estimating the relationship between product characteristics and price.
5.2.4 Price Distribution by Brand
A boxplot visualisation was created to compare price distributions across brands.
top_brands_reg <- names(
sort(table(df_sample$brand), decreasing = TRUE)
)[1:20]
brand_plot <- df_sample %>%
filter(brand %in% top_brands_reg)
ggplot(
brand_plot,
aes(x = reorder(brand, price, median), y = price)
) +
geom_boxplot(fill = "skyblue") +
coord_flip() +
labs(
title = "Price Distribution by Top 20 Brands",
x = "Brand",
y = "Price"
)
Interpretation
The boxplot shows that product prices vary considerably across the top 20 brands. Some brands have higher median prices and a wider spread of prices, while others have lower and more consistent price ranges. Several brands also exhibit extreme high-price outliers, indicating the presence of premium products. These differences suggest that brand has a significant influence on product price and is therefore an important predictor to include in the regression model.
5.3 Train-Test Split
set.seed(123)
train_index <- createDataPartition(
y = df_sample$price,
p = 0.80,
list = FALSE
)
train_data <- df_sample[train_index, ]
test_data <- df_sample[-train_index, ]
dim(train_data)## [1] 118272 8## [1] 29566 8# Ensure train and test use the same factor levels
train_data$brand <- factor(train_data$brand)
train_data$category_code <- factor(train_data$category_code)
test_data$brand <- factor(
test_data$brand,
levels = levels(train_data$brand)
)
test_data$category_code <- factor(
test_data$category_code,
levels = levels(train_data$category_code)
)Interpretation
The sampled dataset was divided into 80% training data and 20% testing data. The training dataset is used to build the regression model, while the testing dataset is used to evaluate its predictive performance on unseen data. Aligning factor levels between the two sets prevents prediction errors when the test set contains brand or category levels absent from training.
5.4 Simple Linear Regression
##
## Call:
## lm(formula = price ~ brand, data = train_data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -706.8 -38.6 -11.7 5.8 5341.7
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 2.70629 19.39700 0.140 0.889039
## brandaccesstyle 28.02514 64.33257 0.436 0.663107
## brandaction 22.73371 163.44203 0.139 0.889377
## brandadidas 39.40371 116.38201 0.339 0.734933
## brandadvantek 11.67260 31.00825 0.376 0.706594
## brandaeg 634.61871 69.03443 9.193 < 2e-16 ***
## brandaerocool 114.86159 30.60403 3.753 0.000175 ***
## brandaimoto 53.40696 32.16629 1.660 0.096849 .
## brandairline 8.97121 31.22200 0.287 0.773855
## brandakku 65.90621 83.42966 0.790 0.429552
## brandaksion 40.23315 22.15294 1.816 0.069349 .
## brandakvafor 6.63411 19.98774 0.332 0.739958
## brandalex 67.64371 163.44203 0.414 0.678970
## brandalfa 7.45371 163.44203 0.046 0.963625
## brandalienware 3617.64371 163.44203 22.134 < 2e-16 ***
## brandalmatv 20.42371 116.38201 0.175 0.860696
## brandaltair 34.30371 163.44203 0.210 0.833759
## brandaltel 44.78843 22.89552 1.956 0.050443 .
## brandaltex 146.65371 46.17410 3.176 0.001493 **
## brandamazon 177.24621 83.42966 2.124 0.033631 *
## brandamd 207.24493 34.26859 6.048 1.47e-09 ***
## brandamigami -0.41629 42.88838 -0.010 0.992256
## brandanymode 20.42371 28.68160 0.712 0.476414
## brandaoc 169.50887 25.45112 6.660 2.75e-11 ***
## brandaoki 119.06756 49.01195 2.429 0.015127 *
## brandapc 21.92471 54.86300 0.400 0.689432
## brandapollo 4.33590 24.44511 0.177 0.859216
## brandapple 625.75330 19.59302 31.938 < 2e-16 ***
## brandaqua 32.58064 49.01195 0.665 0.506212
## brandaquadent 53.47371 57.46811 0.930 0.352117
## brandaquael 35.46371 163.44203 0.217 0.828224
## brandaquapick 78.29371 49.01195 1.597 0.110170
## brandarena 17.87371 163.44203 0.109 0.912919
## brandariston 125.76252 21.19911 5.932 2.99e-09 ***
## brandarktika 39.37121 60.56712 0.650 0.515666
## brandarnica 58.15371 95.68313 0.608 0.543339
## brandart.fit 67.33371 50.70501 1.328 0.184197
## brandartberry -0.50629 163.44203 -0.003 0.997528
## brandartel 130.24371 52.63574 2.474 0.013346 *
## brandasrock 70.92838 46.17410 1.536 0.124515
## brandastonish 13.86371 95.68313 0.145 0.884796
## brandasus 521.30530 20.06904 25.976 < 2e-16 ***
## brandatlant 331.32402 23.99514 13.808 < 2e-16 ***
## brandatmor 40.67205 50.70501 0.802 0.422479
## brandatom 20.42371 163.44203 0.125 0.900555
## brandaudac 20.42371 163.44203 0.125 0.900555
## brandaura -2.54629 116.38201 -0.022 0.982545
## brandausini 6.70518 23.87255 0.281 0.778807
## brandauthor 219.26371 163.44203 1.342 0.179748
## brandava 69.16877 19.52320 3.543 0.000396 ***
## brandavermedia 149.27705 95.68313 1.560 0.118735
## brandavrora -1.03548 28.30256 -0.037 0.970815
## brandawax 2.85371 21.07422 0.135 0.892286
## brandawei 20.04418 31.44413 0.637 0.523831
## brandbaboo 10.42371 116.38201 0.090 0.928633
## brandbabyliss 47.85437 25.73939 1.859 0.063003 .
## brandbaltextile 99.00371 116.38201 0.851 0.394949
## brandbarbie 18.10371 95.68313 0.189 0.849933
## brandbardahl 21.34371 163.44203 0.131 0.896101
## brandbarjher 3.79381 20.91293 0.181 0.856046
## brandbarkan 34.40887 21.85269 1.575 0.115355
## brandbeats 192.70514 64.33257 2.995 0.002741 **
## brandbeeline 38.70371 25.62079 1.511 0.130884
## brandbeko 262.39340 19.86032 13.212 < 2e-16 ***
## brandbelcando 77.43371 163.44203 0.474 0.635666
## brandbelita 0.88371 95.68313 0.009 0.992631
## brandbelitam 0.37371 163.44203 0.002 0.998176
## brandbellissima 41.57355 28.42532 1.463 0.143592
## brandbequiet 60.86371 83.42966 0.730 0.465684
## brandberghoff 12.89333 20.87011 0.618 0.536715
## brandberkley 2.59371 163.44203 0.016 0.987339
## brandberry 34.30371 163.44203 0.210 0.833759
## brandbestway 21.31816 57.46811 0.371 0.710671
## brandbeurer 35.21816 24.52237 1.436 0.150958
## brandbfgoodrich 53.52371 163.44203 0.327 0.743307
## brandbioderma 21.60371 163.44203 0.132 0.894842
## brandbiol 3.32502 39.00427 0.085 0.932065
## brandbiolane 4.23371 163.44203 0.026 0.979334
## brandbiostar 47.26705 95.68313 0.494 0.621310
## brandbirjusa 274.48065 21.84381 12.566 < 2e-16 ***
## brandbisfree 0.97371 163.44203 0.006 0.995247
## brandblackstar 80.60371 116.38201 0.693 0.488575
## brandblehk 28.08371 163.44203 0.172 0.863574
## brandbloody 23.20644 21.43802 1.082 0.279037
## brandbona 77.36271 35.41392 2.185 0.028925 *
## brandborasco -1.92329 41.14725 -0.047 0.962719
## brandbork 316.50962 24.15827 13.102 < 2e-16 ***
## brandborner 55.14371 75.12426 0.734 0.462930
## brandbosch 316.70935 19.82988 15.971 < 2e-16 ***
## brandbose 311.04496 60.56712 5.136 2.82e-07 ***
## brandbradex 9.04091 37.81172 0.239 0.811026
## brandbrateck 136.16371 163.44203 0.833 0.404789
## brandbraun 88.65643 20.07917 4.415 1.01e-05 ***
## brandbrelil 8.15371 163.44203 0.050 0.960212
## brandbridgestone 94.57038 95.68313 0.988 0.322973
## brandbrother 460.23371 69.03443 6.667 2.63e-11 ***
## brandbruder 53.29371 163.44203 0.326 0.744371
## brandbuebchen 1.92371 163.44203 0.012 0.990609
## brandbuff 20.88371 163.44203 0.128 0.898328
## brandbushido 8.14716 35.83878 0.227 0.820169
## brandbyintek 321.34371 163.44203 1.966 0.049289 *
## brandbykski 1.69371 57.46811 0.029 0.976488
## brandcablexpert -1.80629 116.38201 -0.016 0.987617
## brandcalgon 4.91371 163.44203 0.030 0.976016
## brandcamelion -0.38478 20.25732 -0.019 0.984845
## brandcannondale 1038.70371 163.44203 6.355 2.09e-10 ***
## brandcanon 158.57455 21.81754 7.268 3.67e-13 ***
## brandcanyon 95.26705 69.03443 1.380 0.167591
## brandcasada 575.97371 163.44203 3.524 0.000425 ***
## brandcasio 213.58871 116.38201 1.835 0.066473 .
## brandcaso 549.88099 52.63574 10.447 < 2e-16 ***
## brandcaspio 89.86371 64.33257 1.397 0.162458
## brandcatrice -0.63629 163.44203 -0.004 0.996894
## brandcdc 7.38671 54.86300 0.135 0.892897
## brandcelebrat -2.47629 163.44203 -0.015 0.987912
## brandchicco 19.73417 39.66589 0.498 0.618830
## brandchina 5.26669 23.30094 0.226 0.821179
## brandcilek 546.67847 23.67317 23.093 < 2e-16 ***
## brandcliny 0.62371 163.44203 0.004 0.996955
## brandcollecta 0.51371 116.38201 0.004 0.996478
## brandcolorful 42.57735 52.63574 0.809 0.418571
## brandcompliment 0.83231 30.04970 0.028 0.977903
## brandcontinent 23.43475 24.99413 0.938 0.348447
## brandcontinental 147.96371 116.38201 1.271 0.203602
## brandcoolinar 10.38571 24.05902 0.432 0.665977
## brandcort 227.36371 163.44203 1.391 0.164199
## brandcougar 100.52657 47.51275 2.116 0.034366 *
## brandcremesso 8.00352 29.71070 0.269 0.787636
## brandcullmann 46.21157 47.51275 0.973 0.330747
## brandcyberpower 3.05371 163.44203 0.019 0.985093
## brandd-color 13.47371 163.44203 0.082 0.934299
## brandd-link 13.81850 39.00427 0.354 0.723128
## branddaewoo 440.85705 69.03443 6.386 1.71e-10 ***
## branddaikin 575.97371 163.44203 3.524 0.000425 ***
## branddaiwa 38.70371 163.44203 0.237 0.812809
## branddam -2.03629 163.44203 -0.012 0.990060
## branddarina 188.08705 95.68313 1.966 0.049332 *
## branddc-girls 6.53371 50.70501 0.129 0.897471
## branddecoroom 41.36371 163.44203 0.253 0.800208
## branddeepcool 21.31235 25.99093 0.820 0.412222
## branddeerma 36.62371 116.38201 0.315 0.753001
## branddefender 20.42371 69.03443 0.296 0.767347
## branddell 585.54305 46.17410 12.681 < 2e-16 ***
## branddelonghi 239.64988 21.27921 11.262 < 2e-16 ***
## branddelux 7.23630 21.06596 0.344 0.731218
## branddeluxe 33.67093 20.45441 1.646 0.099737 .
## branddemidovskiy 9.53064 49.01195 0.194 0.845820
## branddepileve 1.86371 163.44203 0.011 0.990902
## branddermal -1.77629 95.68313 -0.019 0.985189
## branddermoviva 0.65371 33.28406 0.020 0.984330
## branddesignskin -0.82829 75.12426 -0.011 0.991203
## branddifferent 39.92922 22.41112 1.782 0.074806 .
## branddji 433.99022 26.33548 16.479 < 2e-16 ***
## branddogland 31.93821 21.56896 1.481 0.138676
## branddomini 96.03371 95.68313 1.004 0.315543
## branddougez -0.50629 163.44203 -0.003 0.997528
## branddr.beckmann -0.41629 57.46811 -0.007 0.994220
## brandduracell 0.77240 25.34314 0.030 0.975686
## branddxracer 281.95856 34.26859 8.228 < 2e-16 ***
## branddyson 551.54969 27.24034 20.248 < 2e-16 ***
## brande.gov 11.16371 31.67517 0.352 0.724506
## brandecocool 59.76952 35.01180 1.707 0.087801 .
## brandecologystone 64.32371 83.42966 0.771 0.440712
## brandedifier 12.21149 42.88838 0.285 0.775854
## brandeglo 25.06705 95.68313 0.262 0.793337
## brandegoiste -0.41629 83.42966 -0.005 0.996019
## brandehlektrostandart 22.73371 163.44203 0.139 0.889377
## brandehra 1.00105 31.00825 0.032 0.974246
## brandelari 41.36059 28.06706 1.474 0.140583
## brandelectrolux 397.21067 20.49550 19.380 < 2e-16 ***
## brandemsa 32.63071 41.14725 0.793 0.427766
## brandenergea 23.32848 40.37808 0.578 0.563433
## brandepson 72.10544 20.94406 3.443 0.000576 ***
## brandergolux -0.60616 23.30094 -0.026 0.979246
## brandeset 13.97356 22.62745 0.618 0.536874
## brandetalon 103.75371 163.44203 0.635 0.525557
## brandeuroprint 13.98571 75.12426 0.186 0.852314
## brandeverlast 57.45371 163.44203 0.352 0.725196
## brandeyfel 1.90371 163.44203 0.012 0.990707
## brandezviz 50.74371 83.42966 0.608 0.543042
## brandfender 413.70371 163.44203 2.531 0.011369 *
## brandfiltero 3.16552 26.33548 0.120 0.904325
## brandfissman 11.57348 21.89809 0.529 0.597143
## brandfitokosmetik -1.68629 163.44203 -0.010 0.991768
## brandfixsen 12.92371 163.44203 0.079 0.936975
## brandfizan 13.24371 163.44203 0.081 0.935418
## brandflama 131.13096 35.83878 3.659 0.000253 ***
## brandfly 29.67371 95.68313 0.310 0.756467
## brandforward 194.03371 163.44203 1.187 0.235162
## brandfossil 117.86826 52.63574 2.239 0.025137 *
## brandfranke 117.80705 95.68313 1.231 0.218243
## brandfujifilm 274.31625 28.18319 9.733 < 2e-16 ***
## brandgalaxy 11.93038 95.68313 0.125 0.900772
## brandgamdias 179.14943 64.33257 2.785 0.005358 **
## brandgamemax 47.72731 26.97044 1.770 0.076794 .
## brandgarmin 725.87201 31.91568 22.743 < 2e-16 ***
## brandgefest 278.00538 29.39331 9.458 < 2e-16 ***
## brandgembird 22.73371 163.44203 0.139 0.889377
## brandgenau 339.39871 116.38201 2.916 0.003543 **
## brandgeneris 101.43371 163.44203 0.621 0.534858
## brandgenius 10.76355 20.58809 0.523 0.601111
## brandgerat 5.68390 29.39331 0.193 0.846666
## brandgewa 44.08871 83.42966 0.528 0.597185
## brandgeyzer 1.87371 60.56712 0.031 0.975321
## brandgezatone 43.69064 27.95405 1.563 0.118068
## brandggg 17.93736 21.87065 0.820 0.412128
## brandgiant 547.96871 116.38201 4.708 2.50e-06 ***
## brandgigabyte 240.01371 28.18319 8.516 < 2e-16 ***
## brandgillette 5.66371 83.42966 0.068 0.945876
## brandgiottos 66.71371 163.44203 0.408 0.683143
## brandglasslock 7.16371 27.73687 0.258 0.796196
## brandglassware 15.12858 27.24034 0.555 0.578640
## brandglobal 3.05674 34.26859 0.089 0.928923
## brandgo-sport 778.52371 163.44203 4.763 1.91e-06 ***
## brandgolf -1.54629 163.44203 -0.009 0.992452
## brandgoodride 46.11371 21.27921 2.167 0.030231 *
## brandgoodyear 48.19871 60.56712 0.796 0.426156
## brandgopro 160.28129 34.26859 4.677 2.91e-06 ***
## brandgorenje 308.23955 21.14337 14.579 < 2e-16 ***
## brandgrans 30.74205 28.55161 1.077 0.281608
## brandgreenway 212.50059 30.41270 6.987 2.82e-12 ***
## brandgrillver 62.08371 163.44203 0.380 0.704056
## brandgrohe 160.32706 20.19369 7.939 2.05e-15 ***
## brandgtec -1.43545 50.70501 -0.028 0.977415
## brandgutrend 282.05538 42.88838 6.576 4.84e-11 ***
## brandhabilead 45.65371 23.84292 1.915 0.055524 .
## brandhansa 292.85768 20.72577 14.130 < 2e-16 ***
## brandhaushalt 37.54971 33.59659 1.118 0.263713
## brandhenkel 4.64038 69.03443 0.067 0.946408
## brandherschel 75.62871 116.38201 0.650 0.515802
## brandhintek 54.59705 95.68313 0.571 0.568270
## brandhitachi 2180.14371 116.38201 18.733 < 2e-16 ***
## brandhms 275.04371 163.44203 1.683 0.092412 .
## brandhoco -1.47429 37.81172 -0.039 0.968898
## brandhonor 113.27003 20.52848 5.518 3.44e-08 ***
## brandhotpoint-ariston 305.08151 21.37253 14.274 < 2e-16 ***
## brandhp 297.93821 20.11527 14.812 < 2e-16 ***
## brandhtc 468.33621 83.42966 5.614 1.99e-08 ***
## brandhuawei 181.21265 19.65510 9.220 < 2e-16 ***
## brandhuion 83.38371 69.03443 1.208 0.227105
## brandhuntkey 13.70394 31.22200 0.439 0.660721
## brandhurom 298.19371 116.38201 2.562 0.010402 *
## brandhuter 168.18871 35.41392 4.749 2.04e-06 ***
## brandhygge 631.53371 163.44203 3.864 0.000112 ***
## brandhyperx 70.04210 21.56202 3.248 0.001161 **
## brandhyundai 131.52871 116.38201 1.130 0.258417
## brandid-cooling 6.49943 47.51275 0.137 0.891194
## brandideal 1.20371 32.16629 0.037 0.970149
## brandikins -0.41629 163.44203 -0.003 0.997968
## brandimetec 24.87496 44.97010 0.553 0.580165
## brandincase 94.59010 23.64614 4.000 6.33e-05 ***
## brandindesit 292.47236 20.89949 13.994 < 2e-16 ***
## brandinhouse 12.91660 22.91376 0.564 0.572956
## brandinkax 1.20371 163.44203 0.007 0.994124
## brandinoi 22.73371 163.44203 0.139 0.889377
## brandinsight 11.55371 95.68313 0.121 0.903889
## brandinspector 184.19371 83.42966 2.208 0.027262 *
## brandintel 203.43167 31.22200 6.516 7.27e-11 ***
## brandintex 174.87538 50.70501 3.449 0.000563 ***
## brandionkini 19.44371 163.44203 0.119 0.905304
## brandipower 5.22578 21.88887 0.239 0.811306
## brandiqos 34.06955 38.38777 0.888 0.374806
## brandirbis 221.34571 46.17410 4.794 1.64e-06 ***
## brandivi 6.26654 25.73939 0.243 0.807649
## brandivolia 66.71371 163.44203 0.408 0.683143
## brandiwalk 0.74371 75.12426 0.010 0.992101
## brandjabra 29.53371 40.37808 0.731 0.464518
## brandjaguar 126.52249 30.22804 4.186 2.85e-05 ***
## brandjandeks 120.81976 31.44413 3.842 0.000122 ***
## brandjanome 186.84454 21.61160 8.646 < 2e-16 ***
## brandjbl 46.54844 20.68538 2.250 0.024431 *
## brandjetair 63.82515 24.22685 2.634 0.008428 **
## brandjetpik 75.97371 163.44203 0.465 0.642050
## brandjoby 41.71594 57.46811 0.726 0.467903
## brandjoerex 19.64371 95.68313 0.205 0.837338
## brandjonsbo 19.37371 116.38201 0.166 0.867790
## brandjvc 50.50871 116.38201 0.434 0.664296
## brandkaabo 519.64605 35.41392 14.673 < 2e-16 ***
## brandkama 27.59371 83.42966 0.331 0.740840
## brandkarcher 165.58669 30.60403 5.411 6.29e-08 ***
## brandkaspersky 13.97592 26.12448 0.535 0.592669
## brandkenko 8.84705 95.68313 0.092 0.926331
## brandkenwood 221.15301 22.04499 10.032 < 2e-16 ***
## brandkicx 97.50205 69.03443 1.412 0.157844
## brandking 147.73371 57.46811 2.571 0.010150 *
## brandkingston 47.09641 25.29060 1.862 0.062576 .
## brandkitchenaid 719.05190 52.63574 13.661 < 2e-16 ***
## brandkivi 219.90371 43.88032 5.011 5.41e-07 ***
## brandkmk 1.90371 116.38201 0.016 0.986949
## brandkomax 1.87238 21.44419 0.087 0.930422
## brandkosadaka -1.33629 163.44203 -0.008 0.993477
## brandkramet 6.23934 44.97010 0.139 0.889653
## brandkrups 629.04943 47.51275 13.240 < 2e-16 ***
## brandkumano 4.47371 163.44203 0.027 0.978163
## brandkumho 56.53371 163.44203 0.346 0.729423
## brandkurzweil 1106.29371 163.44203 6.769 1.31e-11 ***
## brandkyocera 113.01371 95.68313 1.181 0.237556
## brandlamart -0.41829 75.12426 -0.006 0.995557
## brandlaurastar 1244.95371 116.38201 10.697 < 2e-16 ***
## brandlavazza 8.42959 23.37007 0.361 0.718324
## brandlego 28.52460 23.59310 1.209 0.226656
## brandlegrand 22.90921 41.14725 0.557 0.577691
## brandlenovo 532.19713 20.15540 26.405 < 2e-16 ***
## brandlenspen 11.16371 116.38201 0.096 0.923582
## brandlg 468.39106 19.65838 23.827 < 2e-16 ***
## brandlihom 148.71260 57.46811 2.588 0.009662 **
## brandlion -0.69629 163.44203 -0.004 0.996601
## brandloewe 4626.90371 163.44203 28.309 < 2e-16 ***
## brandlogitech 27.50277 20.33329 1.353 0.176187
## brandlol 4.21371 163.44203 0.026 0.979432
## brandlori -0.86629 116.38201 -0.007 0.994061
## brandlotte 697.50371 116.38201 5.993 2.06e-09 ***
## brandlowepro 25.04371 95.68313 0.262 0.793525
## brandluch -1.73629 163.44203 -0.011 0.991524
## brandlumax 15.56157 47.51275 0.328 0.743272
## brandluminarc 75.12958 35.83878 2.096 0.036056 *
## brandluxell 28.13705 95.68313 0.294 0.768709
## brandmaestro 15.12581 21.41371 0.706 0.479965
## brandmakita 353.63371 83.42966 4.239 2.25e-05 ***
## brandmanfrotto 60.31826 52.63574 1.146 0.251816
## brandmarcato 83.12154 39.00427 2.131 0.033084 *
## brandmarcel 0.51371 163.44203 0.003 0.997492
## brandmarley 66.71871 116.38201 0.573 0.566461
## brandmarshall 143.87871 69.03443 2.084 0.037148 *
## brandmart 182.40295 37.27207 4.894 9.90e-07 ***
## brandmarvel 55.14371 163.44203 0.337 0.735823
## brandmatrix 8.53705 95.68313 0.089 0.928906
## brandmattel 1.71998 23.12792 0.074 0.940718
## brandmaxwell 16.72067 20.86057 0.802 0.422818
## brandmcdavid 9.88371 116.38201 0.085 0.932321
## brandmedisana 50.50871 116.38201 0.434 0.664296
## brandmegogo 6.26080 22.95084 0.273 0.785014
## brandmercusys 15.80859 32.42765 0.488 0.625903
## brandmetabo 44.57038 95.68313 0.466 0.641351
## brandmetalions 16.71371 116.38201 0.144 0.885808
## brandmichelin 123.42371 163.44203 0.755 0.450159
## brandmicro 228.75371 163.44203 1.400 0.161635
## brandmicrolab 29.19607 22.89552 1.275 0.202245
## brandmicrosoft 109.28416 22.87746 4.777 1.78e-06 ***
## brandmidea 174.62724 21.10810 8.273 < 2e-16 ***
## brandmilight 10.62038 95.68313 0.111 0.911620
## brandmirage 43.56371 163.44203 0.267 0.789824
## brandmisty 221.67371 44.97010 4.929 8.26e-07 ***
## brandmlife 4.21371 50.70501 0.083 0.933770
## brandmonge 10.58371 163.44203 0.065 0.948369
## brandmonkart 26.43371 163.44203 0.162 0.871518
## brandmoshi 28.10371 37.81172 0.743 0.457329
## brandmotorola 47.98371 163.44203 0.294 0.769077
## brandmoulinex 101.66436 20.14045 5.048 4.48e-07 ***
## brandmoxom 8.84371 69.03443 0.128 0.898065
## brandmsi 391.56553 52.63574 7.439 1.02e-13 ***
## brandmueller 0.51371 163.44203 0.003 0.997492
## brandmujjo 0.92371 41.98100 0.022 0.982445
## brandmuljhtidom 0.84402 20.82964 0.041 0.967678
## brandnavien 637.89371 60.56712 10.532 < 2e-16 ***
## brandnavitel 77.13371 163.44203 0.472 0.636975
## brandneo 34.54794 19.73741 1.750 0.080056 .
## brandneoline 138.49275 28.30256 4.893 9.93e-07 ***
## brandneptun 31.51705 95.68313 0.329 0.741862
## brandnika 20.83709 21.63355 0.963 0.335457
## brandnikon 773.50467 40.37808 19.157 < 2e-16 ***
## brandninebot 460.32955 38.38777 11.992 < 2e-16 ***
## brandnintendo 181.23097 29.87736 6.066 1.32e-09 ***
## brandnivea 0.14371 33.92437 0.004 0.996620
## brandnokia 39.96841 22.35826 1.788 0.073837 .
## brandnokian 42.87371 163.44203 0.262 0.793077
## brandnommi 140.79371 163.44203 0.861 0.389004
## brandnone 44.74187 20.49196 2.183 0.029009 *
## brandnordland 16.72934 44.97010 0.372 0.709886
## brandnovatrack 167.40971 75.12426 2.228 0.025853 *
## brandnuk 10.60371 163.44203 0.065 0.948272
## brandnv-print 12.35705 50.70501 0.244 0.807460
## brandnzxt 187.60621 83.42966 2.249 0.024535 *
## brandockel 124.58371 163.44203 0.762 0.445912
## brandokko 13.07775 23.46617 0.557 0.577322
## brandokuma 42.79371 95.68313 0.447 0.654700
## brandolimpik -1.12586 30.60403 -0.037 0.970654
## brandolympus 82.92371 163.44203 0.507 0.611904
## brandomron 46.77284 30.80241 1.518 0.128896
## brandoppo 195.91449 19.77705 9.906 < 2e-16 ***
## brandoptoma 552.82371 163.44203 3.382 0.000719 ***
## brandoral-b 27.21466 26.19335 1.039 0.298811
## brandorgan -1.25709 23.49093 -0.054 0.957323
## brandorico 13.77514 64.33257 0.214 0.830451
## brandosz 21.87401 27.73687 0.789 0.430332
## brandowner -1.33629 163.44203 -0.008 0.993477
## brandozone 7.05734 27.53071 0.256 0.797686
## brandpaclan 8.67121 83.42966 0.104 0.917222
## brandpalisad 4.79705 69.03443 0.069 0.944601
## brandpalit 250.20371 69.03443 3.624 0.000290 ***
## brandpanasonic 66.90055 20.27970 3.299 0.000971 ***
## brandpaperline 8.84371 163.44203 0.054 0.956848
## brandpasabahce 0.96145 20.69276 0.046 0.962941
## brandpastel 44.17353 24.76922 1.783 0.074523 .
## brandpatriot 48.20201 31.91568 1.510 0.130972
## brandpccooler 9.19371 116.38201 0.079 0.937036
## brandpemco 6.17871 116.38201 0.053 0.957660
## brandperilla 33.12550 21.68651 1.527 0.126647
## brandpeterhof 6.81577 22.06589 0.309 0.757411
## brandpetzl 37.54871 116.38201 0.323 0.746974
## brandpgytech 8.07371 95.68313 0.084 0.932755
## brandphantom 96.80371 163.44203 0.592 0.553663
## brandphilips 78.56929 19.68125 3.992 6.55e-05 ***
## brandpixel 136.16371 116.38201 1.170 0.242015
## brandplantronics 32.24615 31.91568 1.010 0.312328
## brandplayme 275.04371 163.44203 1.683 0.092412 .
## brandplextor 55.83371 163.44203 0.342 0.732644
## brandpocketbook 177.68371 32.42765 5.479 4.28e-08 ***
## brandpolaris 36.28730 19.76590 1.836 0.066382 .
## brandpolaroid 0.04871 83.42966 0.001 0.999534
## brandpolimerbiht 0.44796 20.10849 0.022 0.982227
## brandportcase 0.69371 42.88838 0.016 0.987095
## brandpowerplant 30.46571 75.12426 0.406 0.685083
## brandpowertrac 33.78621 60.56712 0.558 0.576961
## brandpozis 336.42161 26.88464 12.514 < 2e-16 ***
## brandpresident 334.07371 163.44203 2.044 0.040957 *
## brandprestigio 61.18613 35.83878 1.707 0.087776 .
## brandprocab 6.53371 163.44203 0.040 0.968113
## brandpromtorgservis 25.04371 163.44203 0.153 0.878220
## brandproscreen 177.36371 163.44203 1.085 0.277845
## brandprovence 1.85712 21.23829 0.087 0.930320
## brandpyrex 14.24826 24.37017 0.585 0.558778
## brandrapoo 36.15324 31.67517 1.141 0.253716
## brandrastar 27.17191 27.24034 0.997 0.318530
## brandrazer 130.90557 31.44413 4.163 3.14e-05 ***
## brandredmond 44.12845 20.32816 2.171 0.029948 *
## brandregnum 74.08281 34.26859 2.162 0.030633 *
## brandremington 39.13142 22.59682 1.732 0.083326 .
## brandresanta 52.52800 47.51275 1.106 0.268921
## brandresto 14.82705 69.03443 0.215 0.829941
## brandrioba 0.97371 163.44203 0.006 0.995247
## brandritmix 6.20086 40.37808 0.154 0.877949
## brandriva 25.04871 116.38201 0.215 0.829590
## brandrivacase 26.67320 28.81546 0.926 0.354627
## brandroadx 41.64038 95.68313 0.435 0.663425
## brandrockstar 57.45371 35.83878 1.603 0.108912
## brandrondell 23.17180 20.15287 1.150 0.250228
## brandrosneftjh 1.67371 163.44203 0.010 0.991829
## brandrossija 14.17038 95.68313 0.148 0.882267
## brandrowenta 39.77443 20.17631 1.971 0.048687 *
## brandruggear 52.59371 69.03443 0.762 0.446153
## brandsakura -1.57629 163.44203 -0.010 0.992305
## brandsamsonite 112.35229 64.33257 1.746 0.080739 .
## brandsamsung 247.30580 19.43897 12.722 < 2e-16 ***
## brandsamura 24.58371 163.44203 0.150 0.880439
## brandsanc 207.74408 36.76534 5.651 1.60e-08 ***
## brandsaramonic 67.37657 64.33257 1.047 0.294956
## brandsatechi 80.48746 60.56712 1.329 0.183884
## brandsavic 14.80871 83.42966 0.177 0.859116
## brandsbs 3.44407 29.09550 0.118 0.905774
## brandscarlett 20.01984 19.97912 1.002 0.316327
## brandschwiizer 18.10371 163.44203 0.111 0.911803
## brandscreentec -2.47629 60.56712 -0.041 0.967388
## brandseagate 87.55038 95.68313 0.915 0.360192
## brandseasonic 258.84371 163.44203 1.584 0.113264
## brandselect -1.57629 163.44203 -0.010 0.992305
## brandsencor 50.53554 23.41754 2.158 0.030928 *
## brandsennheiser 210.23371 163.44203 1.286 0.198345
## brandshelkovica 11.18371 163.44203 0.068 0.945446
## brandship 6.28945 22.15294 0.284 0.776479
## brandsibrtekh -1.10629 163.44203 -0.007 0.994599
## brandsigma 344.49371 163.44203 2.108 0.035055 *
## brandsimax 9.29259 22.46591 0.414 0.679145
## brandsimfer 121.83038 31.67517 3.846 0.000120 ***
## brandsinger 160.80136 43.88032 3.665 0.000248 ***
## brandsjcam 113.78205 69.03443 1.648 0.099316 .
## brandskullcandy 8.84371 163.44203 0.054 0.956848
## brandsmart 9.54371 163.44203 0.058 0.953436
## brandsmeg 326.74705 95.68313 3.415 0.000638 ***
## brandsmile 501.25371 163.44203 3.067 0.002164 **
## brandsony 211.10402 19.94483 10.584 < 2e-16 ***
## brandsparta 0.51371 163.44203 0.003 0.997492
## brandspiegelau 20.47721 41.14725 0.498 0.618727
## brandspigen -2.14471 41.98100 -0.051 0.959256
## brandsportop 460.23371 163.44203 2.816 0.004865 **
## brandsports 719.49571 75.12426 9.577 < 2e-16 ***
## brandstaedtler -1.33629 163.44203 -0.008 0.993477
## brandstarline 127.64962 39.66589 3.218 0.001291 **
## brandstatus 12.54571 75.12426 0.167 0.867371
## brandstaub 170.65171 75.12426 2.272 0.023113 *
## brandsteelseries 92.32657 36.28847 2.544 0.010953 *
## brandstels 249.35371 163.44203 1.526 0.127102
## brandsumdex 22.36429 29.71070 0.753 0.451610
## brandsuperlux 92.17371 69.03443 1.335 0.181819
## brandsvc 26.83614 22.39773 1.198 0.230856
## brandsvetocopy 1.46371 31.00825 0.047 0.962351
## brandsynology 606.09371 163.44203 3.708 0.000209 ***
## brandtacx 830.60371 163.44203 5.082 3.74e-07 ***
## brandtailg 493.61657 47.51275 10.389 < 2e-16 ***
## brandtamron 830.60371 75.12426 11.056 < 2e-16 ***
## brandtaoran 251.90371 163.44203 1.541 0.123261
## brandtayfun 2.63271 32.16629 0.082 0.934769
## brandtcl 249.45962 22.02441 11.327 < 2e-16 ***
## brandtechnodom 9.88419 20.09103 0.492 0.622741
## brandtechnogym 4491.71371 163.44203 27.482 < 2e-16 ***
## brandtechnomax 109.30371 163.44203 0.669 0.503649
## brandtefal 77.00286 19.58151 3.932 8.41e-05 ***
## brandtemdan -2.46429 75.12426 -0.033 0.973832
## brandtenda 6.53371 163.44203 0.040 0.968113
## brandthefaceshop 3.98371 21.77501 0.183 0.854838
## brandthermaltake 61.48746 38.38777 1.602 0.109214
## brandthermex 104.54697 22.49406 4.648 3.36e-06 ***
## brandthomas 198.44025 26.97044 7.358 1.88e-13 ***
## brandthule 87.54371 163.44203 0.536 0.592218
## brandtigar 87.54371 163.44203 0.536 0.592218
## brandtigernu 43.56371 163.44203 0.267 0.789824
## brandtimberk 60.92871 116.38201 0.524 0.600611
## brandtimson 6.74903 34.63057 0.195 0.845482
## brandtion 582.92371 163.44203 3.567 0.000362 ***
## brandtoday -0.41629 163.44203 -0.003 0.997968
## brandtopperr 11.80524 30.80241 0.383 0.701530
## brandtornado 31.98871 83.42966 0.383 0.701408
## brandtoro 0.99652 21.25342 0.047 0.962603
## brandtoshiba 40.12050 36.28847 1.106 0.268902
## brandtoyo 78.28871 116.38201 0.673 0.501147
## brandtp-link 28.94911 20.39236 1.420 0.155725
## brandtramp 17.20943 64.33257 0.268 0.789079
## brandtranscend 25.56209 20.21679 1.264 0.206089
## brandtribe 316.71371 75.12426 4.216 2.49e-05 ***
## brandtrio 0.47104 27.33474 0.017 0.986251
## brandtrust 27.93742 23.39366 1.194 0.232390
## brandtucano 32.20771 37.81172 0.852 0.394331
## brandturbo -1.57295 95.68313 -0.016 0.986884
## brandtvs 27.35345 20.13479 1.359 0.174302
## brandtyr 23.89371 163.44203 0.146 0.883771
## brandubisoft 44.72371 116.38201 0.384 0.700769
## brandUnknown 9.01750 19.42622 0.464 0.642511
## brandurbanears 159.30371 163.44203 0.975 0.329721
## branduriage 11.02371 163.44203 0.067 0.946226
## brandusams 5.21038 35.41392 0.147 0.883031
## branduteki 2.60676 30.80241 0.085 0.932557
## brandvarta 1.85437 21.25342 0.087 0.930472
## brandventa 280.83871 116.38201 2.413 0.015820 *
## brandviatti 49.54538 69.03443 0.718 0.472949
## brandvictoria 22.70874 22.87746 0.993 0.320895
## brandvikhrjh 57.93205 33.28406 1.741 0.081768 .
## brandvirtuix 12.31371 163.44203 0.075 0.939944
## brandvitehks -0.46829 54.86300 -0.009 0.993190
## brandvitek 24.62175 19.84326 1.241 0.214678
## brandvivo 163.45964 23.32371 7.008 2.43e-12 ***
## brandvoin 20.42371 163.44203 0.125 0.900555
## brandvortex 2.95371 116.38201 0.025 0.979752
## brandwacom 161.95443 47.51275 3.409 0.000653 ***
## brandweber 14.45871 83.42966 0.173 0.862413
## brandwethepeople 483.38371 163.44203 2.958 0.003102 **
## brandwhirlpool 1056.29871 83.42966 12.661 < 2e-16 ***
## brandwilmax 7.71071 25.29060 0.305 0.760455
## brandwilson 10.92371 163.44203 0.067 0.946713
## brandwintek 29.38871 60.56712 0.485 0.627517
## brandwmf 307.22371 75.12426 4.090 4.33e-05 ***
## brandwonlex 36.62371 163.44203 0.224 0.822697
## brandwxd 8.84371 116.38201 0.076 0.939428
## brandx-game 19.21059 21.22340 0.905 0.365382
## brandxbox 310.56201 31.91568 9.731 < 2e-16 ***
## brandxerox 155.14121 83.42966 1.860 0.062952 .
## brandxiaomi 131.58192 19.84741 6.630 3.38e-11 ***
## brandxp-pen 165.53910 49.01195 3.378 0.000732 ***
## brandyokohama 82.92038 95.68313 0.867 0.386155
## brandyonker 64.40371 163.44203 0.394 0.693548
## brandyoobao 24.23141 32.42765 0.747 0.454917
## brandzala -0.64629 83.42966 -0.008 0.993819
## brandzalman 96.34371 116.38201 0.828 0.407772
## brandzeppelin 23.66371 163.44203 0.145 0.884882
## brandzhiyun 206.56788 50.70501 4.074 4.63e-05 ***
## brandzhorka -1.18629 116.38201 -0.010 0.991867
## brandzowie 82.48371 44.97010 1.834 0.066628 .
## brandzugo 25.07371 75.12426 0.334 0.738559
## brandzwilling 83.68018 43.88032 1.907 0.056522 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 162.3 on 117704 degrees of freedom
## Multiple R-squared: 0.4836, Adjusted R-squared: 0.4812
## F-statistic: 194.4 on 567 and 117704 DF, p-value: < 2.2e-16Interpretation
A Simple Linear Regression model was developed using brand as the predictor variable and price as the response variable. This model serves as a baseline to examine the influence of brand on product price before incorporating additional predictors in the Multiple Linear Regression model.
5.5 Multiple Linear Regression
##
## Call:
## lm(formula = price ~ brand + category_code, data = train_data)
##
## Residuals:
## Min 1Q Median 3Q Max
## -818.8 -33.3 -0.1 7.7 5033.4
##
## Coefficients: (3 not defined because of singularities)
## Estimate Std. Error
## (Intercept) -3.884e+01 2.070e+01
## brandaccesstyle 2.803e+01 5.638e+01
## brandaction 2.273e+01 1.432e+02
## brandadidas 3.940e+01 1.020e+02
## brandadvantek 1.167e+01 2.718e+01
## brandaeg 5.376e+02 6.064e+01
## brandaerocool 3.817e+01 3.378e+01
## brandaimoto 4.327e+01 2.891e+01
## brandairline 8.480e+00 2.753e+01
## brandakku 6.591e+01 7.312e+01
## brandaksion 2.728e+01 1.972e+01
## brandakvafor 6.634e+00 1.752e+01
## brandalex 6.764e+01 1.432e+02
## brandalfa 1.423e+01 1.850e+02
## brandalienware 3.618e+03 1.432e+02
## brandalmatv 2.042e+01 1.020e+02
## brandaltair 3.430e+01 1.432e+02
## brandaltel 1.327e+02 2.954e+01
## brandaltex 1.941e+02 4.172e+01
## brandamazon 1.013e+02 8.428e+01
## brandamd 8.717e+01 3.155e+01
## brandamigami -4.163e-01 3.759e+01
## brandanymode 3.502e+01 2.628e+01
## brandaoc -1.570e+02 2.256e+01
## brandaoki 7.626e+01 4.836e+01
## brandapc -2.267e+02 4.825e+01
## brandapollo 6.965e+00 2.144e+01
## brandapple 5.332e+02 1.730e+01
## brandaqua -1.367e+02 4.549e+01
## brandaquadent 5.347e+01 5.037e+01
## brandaquael 3.546e+01 1.432e+02
## brandaquapick 7.829e+01 4.295e+01
## brandarena 1.787e+01 1.432e+02
## brandariston -4.356e+01 2.385e+01
## brandarktika 4.491e+01 5.312e+01
## brandarnica 3.807e+01 8.392e+01
## brandart.fit 2.137e+01 4.784e+01
## brandartberry -5.063e-01 1.432e+02
## brandartel 1.302e+02 4.613e+01
## brandasrock 1.251e+02 4.819e+01
## brandastonish 1.386e+01 8.386e+01
## brandasus 2.006e+02 1.842e+01
## brandatlant -5.883e+01 2.132e+01
## brandatmor -1.287e+02 4.689e+01
## brandatom 2.042e+01 1.432e+02
## brandaudac -4.249e+01 1.440e+02
## brandaura -2.546e+00 1.020e+02
## brandausini 6.705e+00 2.092e+01
## brandauthor 1.920e+02 1.521e+02
## brandava 6.388e+00 1.717e+01
## brandavermedia 1.576e+02 8.888e+01
## brandavrora -1.035e+00 2.480e+01
## brandawax -1.364e+02 1.862e+01
## brandawei 8.157e+01 2.764e+01
## brandbaboo 2.042e+01 1.432e+02
## brandbabyliss 5.216e+01 2.265e+01
## brandbaltextile 9.900e+01 1.020e+02
## brandbarbie -1.858e+01 9.293e+01
## brandbardahl 2.134e+01 1.432e+02
## brandbarjher 3.794e+00 1.833e+01
## brandbarkan 3.441e+01 1.915e+01
## brandbeats 2.767e+02 5.645e+01
## brandbeeline 1.266e+02 3.121e+01
## brandbeko 3.706e+01 1.763e+01
## brandbelcando -2.764e+01 1.460e+02
## brandbelita -1.241e+01 8.426e+01
## brandbelitam 3.737e-01 1.432e+02
## brandbellissima 4.259e+01 2.501e+01
## brandbequiet 2.819e+02 7.391e+01
## brandberghoff 2.195e+01 1.835e+01
## brandberkley 2.594e+00 1.432e+02
## brandberry 3.430e+01 1.432e+02
## brandbestway 2.132e+01 5.037e+01
## brandbeurer 4.137e+01 2.200e+01
## brandbfgoodrich 5.352e+01 1.432e+02
## brandbioderma 2.160e+01 1.432e+02
## brandbiol 4.555e+00 3.426e+01
## brandbiolane 4.822e+01 1.433e+02
## brandbiostar 1.015e+02 8.784e+01
## brandbirjusa -1.184e+02 1.947e+01
## brandbisfree 9.737e-01 1.432e+02
## brandblackstar 1.786e+02 1.021e+02
## brandblehk 4.211e+01 1.433e+02
## brandbloody 7.969e+01 1.918e+01
## brandbona 7.736e+01 3.104e+01
## brandborasco -1.923e+00 3.606e+01
## brandbork 3.016e+02 2.128e+01
## brandborner 7.293e+01 6.591e+01
## brandbosch 2.043e+02 1.751e+01
## brandbose 3.723e+02 5.312e+01
## brandbradex 9.930e+00 3.319e+01
## brandbrateck 1.362e+02 1.432e+02
## brandbraun 8.778e+01 1.774e+01
## brandbrelil 8.154e+00 1.432e+02
## brandbridgestone 9.457e+01 8.386e+01
## brandbrother 5.297e+02 6.136e+01
## brandbruder 5.329e+01 1.432e+02
## brandbuebchen 1.924e+00 1.432e+02
## brandbuff 2.088e+01 1.432e+02
## brandbushido 8.147e+00 3.141e+01
## brandbyintek 2.708e+02 1.604e+02
## brandbykski 2.626e+01 5.038e+01
## brandcablexpert 6.472e+00 1.062e+02
## brandcalgon 4.914e+00 1.432e+02
## brandcamelion -8.709e+01 2.193e+01
## brandcannondale 1.011e+03 1.521e+02
## brandcanon 3.263e+02 2.094e+01
## brandcanyon -1.965e+02 6.071e+01
## brandcasada 5.804e+02 1.436e+02
## brandcasio -7.286e+01 1.245e+02
## brandcaso 1.570e+02 4.627e+01
## brandcaspio 8.986e+01 5.638e+01
## brandcatrice -6.363e-01 1.432e+02
## brandcdc 7.387e+00 4.808e+01
## brandcelebrat 9.551e+01 1.433e+02
## brandchicco 1.616e+01 3.664e+01
## brandchina -3.499e+02 2.119e+01
## brandcilek 5.469e+02 2.075e+01
## brandcliny -2.546e+02 1.438e+02
## brandcollecta 5.137e-01 1.020e+02
## brandcolorful 4.308e+01 4.676e+01
## brandcompliment -8.384e+00 2.634e+01
## brandcontinent -7.086e+01 2.322e+01
## brandcontinental 1.480e+02 1.020e+02
## brandcoolinar 1.664e+01 2.115e+01
## brandcort 2.274e+02 1.432e+02
## brandcougar 5.789e+01 4.384e+01
## brandcremesso 8.429e+00 2.604e+01
## brandcullmann -1.235e+02 4.186e+01
## brandcyberpower -3.521e+02 1.434e+02
## brandd-color 1.347e+01 1.432e+02
## brandd-link 1.017e+02 4.047e+01
## branddaewoo 8.882e+01 6.056e+01
## branddaikin 3.655e+02 1.433e+02
## branddaiwa 3.870e+01 1.432e+02
## branddam -2.036e+00 1.432e+02
## branddarina 1.881e+02 8.386e+01
## branddc-girls -3.015e+01 5.982e+01
## branddecoroom 6.849e+01 1.700e+02
## branddeepcool -1.341e+02 2.381e+01
## branddeerma 3.662e+01 1.020e+02
## branddefender 1.679e+01 6.226e+01
## branddell 2.570e+02 4.082e+01
## branddelonghi 2.257e+02 1.884e+01
## branddelux 6.211e+01 1.900e+01
## branddeluxe 3.587e+01 1.793e+01
## branddemidovskiy 6.139e+00 4.355e+01
## branddepileve 1.864e+00 1.432e+02
## branddermal -1.776e+00 8.386e+01
## branddermoviva 6.537e-01 2.917e+01
## branddesignskin -8.283e-01 6.584e+01
## branddifferent -1.247e+02 2.026e+01
## branddji 4.272e+02 2.309e+01
## branddogland 7.190e+01 2.441e+01
## branddomini 1.008e+02 8.453e+01
## branddougez 7.772e+00 1.462e+02
## branddr.beckmann -4.163e-01 5.037e+01
## brandduracell -8.093e+01 2.655e+01
## branddxracer 1.841e+02 4.003e+01
## branddyson 5.432e+02 2.391e+01
## brande.gov 1.116e+01 2.776e+01
## brandecocool 5.977e+01 3.068e+01
## brandecologystone 7.111e+01 7.661e+01
## brandedifier 1.102e+02 3.774e+01
## brandeglo 3.386e+01 1.020e+02
## brandegoiste -4.163e-01 7.312e+01
## brandehlektrostandart 2.273e+01 1.432e+02
## brandehra -1.057e+02 2.850e+01
## brandelari 3.291e+01 2.540e+01
## brandelectrolux 2.837e+02 1.812e+01
## brandemsa 3.263e+01 3.606e+01
## brandenergea -7.844e+00 3.550e+01
## brandepson 2.994e+02 2.087e+01
## brandergolux -1.207e+02 2.259e+01
## brandeset -2.091e+02 2.004e+01
## brandetalon -6.557e+01 1.440e+02
## brandeuroprint 3.054e+02 6.686e+01
## brandeverlast 5.745e+01 1.432e+02
## brandeyfel -1.360e+02 1.528e+02
## brandezviz 5.074e+01 7.312e+01
## brandfender 4.137e+02 1.432e+02
## brandfiltero -2.883e+00 2.310e+01
## brandfissman 1.693e+01 1.926e+01
## brandfitokosmetik -1.686e+00 1.432e+02
## brandfixsen 1.292e+01 1.432e+02
## brandfizan 1.324e+01 1.432e+02
## brandflama 1.311e+02 3.141e+01
## brandfly -1.095e+02 8.389e+01
## brandforward 1.940e+02 1.432e+02
## brandfossil 1.179e+02 4.613e+01
## brandfranke 1.359e+02 1.037e+02
## brandfujifilm 2.458e+02 2.495e+01
## brandgalaxy 2.688e+01 8.396e+01
## brandgamdias 7.407e+01 6.303e+01
## brandgamemax 3.761e+01 2.450e+01
## brandgarmin 7.162e+02 2.863e+01
## brandgefest 2.767e+02 2.576e+01
## brandgembird 3.347e+01 1.467e+02
## brandgenau 2.705e+02 1.054e+02
## brandgeneris 1.014e+02 1.432e+02
## brandgenius 3.030e+01 1.864e+01
## brandgerat 5.961e+00 2.576e+01
## brandgewa 4.409e+01 7.312e+01
## brandgeyzer 1.874e+00 5.308e+01
## brandgezatone 4.656e+01 2.527e+01
## brandggg 1.313e+01 2.130e+01
## brandgiant 5.207e+02 1.141e+02
## brandgigabyte 1.962e+02 3.096e+01
## brandgillette 5.664e+00 7.312e+01
## brandgiottos -2.949e+02 1.433e+02
## brandglasslock 7.164e+00 2.431e+01
## brandglassware 1.609e+01 2.404e+01
## brandglobal 1.010e+02 3.022e+01
## brandgo-sport 7.785e+02 1.432e+02
## brandgolf -1.546e+00 1.432e+02
## brandgoodride 4.611e+01 1.865e+01
## brandgoodyear -5.711e+01 6.525e+01
## brandgopro 1.603e+02 3.003e+01
## brandgorenje 5.997e+01 1.871e+01
## brandgrans 3.865e+01 2.540e+01
## brandgreenway 1.875e+02 5.396e+01
## brandgrillver 6.208e+01 1.432e+02
## brandgrohe 1.603e+02 1.770e+01
## brandgtec 6.546e+00 4.444e+01
## brandgutrend 2.620e+02 3.772e+01
## brandhabilead 4.565e+01 2.090e+01
## brandhansa 1.966e+02 1.832e+01
## brandhaushalt 1.059e+01 3.188e+01
## brandhenkel 4.640e+00 6.050e+01
## brandherschel 8.670e+01 1.021e+02
## brandhintek 4.996e+01 1.129e+02
## brandhitachi 1.787e+03 1.021e+02
## brandhms 1.372e+02 1.528e+02
## brandhoco -1.474e+00 3.314e+01
## brandhonor -9.319e+00 1.812e+01
## brandhotpoint-ariston 9.134e+01 1.890e+01
## brandhp 2.745e+02 1.865e+01
## brandhtc 4.683e+02 7.312e+01
## brandhuawei 7.226e+01 1.736e+01
## brandhuion -7.475e+00 6.075e+01
## brandhuntkey -3.001e+02 2.800e+01
## brandhurom 2.569e+02 1.025e+02
## brandhuter 1.330e+02 4.179e+01
## brandhygge 6.315e+02 1.432e+02
## brandhyperx 1.217e+02 1.926e+01
## brandhyundai -3.780e+01 1.031e+02
## brandid-cooling 1.539e+02 4.412e+01
## brandideal 1.204e+00 2.819e+01
## brandikins -4.163e-01 1.432e+02
## brandimetec 3.772e+01 3.957e+01
## brandincase 9.459e+01 2.072e+01
## brandindesit -3.329e+01 1.855e+01
## brandinhouse 1.588e+01 2.011e+01
## brandinkax 9.698e+01 1.434e+02
## brandinoi 7.022e+01 1.436e+02
## brandinsight -5.860e+01 8.387e+01
## brandinspector 2.063e+02 7.499e+01
## brandintel 8.642e+01 2.895e+01
## brandintex 1.749e+02 4.444e+01
## brandionkini 1.944e+01 1.432e+02
## brandipower -3.500e+02 2.000e+01
## brandiqos 3.407e+01 3.364e+01
## brandirbis -1.339e+02 4.086e+01
## brandivi -3.553e+02 2.283e+01
## brandivolia 7.025e+01 1.434e+02
## brandiwalk 7.437e-01 6.584e+01
## brandjabra 5.286e+01 3.540e+01
## brandjaguar 9.682e+01 3.065e+01
## brandjandeks 1.316e+02 2.842e+01
## brandjanome 1.360e+02 2.761e+01
## brandjbl 1.200e+02 1.831e+01
## brandjetair -5.075e+00 2.172e+01
## brandjetpik 7.597e+01 1.432e+02
## brandjoby -2.354e+02 5.070e+01
## brandjoerex -3.670e+02 8.670e+01
## brandjonsbo 1.340e+02 1.032e+02
## brandjvc 9.205e+01 1.027e+02
## brandkaabo 1.330e+02 3.806e+01
## brandkama 2.759e+01 7.312e+01
## brandkarcher 1.562e+02 2.741e+01
## brandkaspersky -1.880e+02 2.304e+01
## brandkenko -2.479e+02 8.957e+01
## brandkenwood 1.809e+02 1.964e+01
## brandkicx -6.282e+01 8.447e+01
## brandking 1.477e+02 5.037e+01
## brandkingston 5.060e+01 2.509e+01
## brandkitchenaid 6.475e+02 4.631e+01
## brandkivi -1.417e+02 3.862e+01
## brandkmk 1.904e+00 1.020e+02
## brandkomax 1.872e+00 1.879e+01
## brandkosadaka -1.336e+00 1.432e+02
## brandkramet -3.867e+02 3.957e+01
## brandkrups 6.354e+02 4.177e+01
## brandkumano -3.571e+02 1.433e+02
## brandkumho 5.653e+01 1.432e+02
## brandkurzweil 8.198e+02 1.600e+02
## brandkyocera 2.386e+02 8.449e+01
## brandlamart 2.551e+00 6.585e+01
## brandlaurastar 1.226e+03 1.021e+02
## brandlavazza 8.430e+00 2.048e+01
## brandlego 2.852e+01 2.068e+01
## brandlegrand -3.323e+02 3.651e+01
## brandlenovo 2.013e+02 1.843e+01
## brandlenspen 1.116e+01 1.020e+02
## brandlg 1.922e+02 1.740e+01
## brandlihom 1.789e+02 5.167e+01
## brandlion -1.952e+02 1.440e+02
## brandloewe 4.265e+03 1.433e+02
## brandlogitech 6.359e+01 1.850e+01
## brandlol -3.247e+01 1.487e+02
## brandlori -8.663e-01 1.020e+02
## brandlotte 6.831e+02 1.525e+02
## brandlowepro 3.981e+01 8.401e+01
## brandluch -1.736e+00 1.432e+02
## brandlumax 1.556e+01 4.164e+01
## brandluminarc 7.513e+01 3.141e+01
## brandluxell 1.165e+01 8.426e+01
## brandmaestro 1.931e+01 1.880e+01
## brandmakita 3.566e+02 7.952e+01
## brandmanfrotto -5.909e+01 4.634e+01
## brandmarcato 8.335e+01 3.419e+01
## brandmarcel 5.137e-01 1.432e+02
## brandmarley 1.157e+02 1.020e+02
## brandmarshall 2.255e+02 6.057e+01
## brandmart 2.299e+02 3.420e+01
## brandmarvel 5.514e+01 1.432e+02
## brandmatrix 8.537e+00 8.386e+01
## brandmattel 4.003e+01 2.523e+01
## brandmaxwell 2.243e+01 1.840e+01
## brandmcdavid 9.884e+00 1.020e+02
## brandmedisana 5.496e+01 1.024e+02
## brandmegogo -3.553e+02 2.042e+01
## brandmercusys 1.037e+02 3.574e+01
## brandmetabo 4.457e+01 8.386e+01
## brandmetalions 1.671e+01 1.020e+02
## brandmichelin 1.234e+02 1.432e+02
## brandmicro -1.579e+02 1.449e+02
## brandmicrolab -2.196e+02 2.057e+01
## brandmicrosoft -1.729e+02 2.051e+01
## brandmidea -9.722e+01 1.926e+01
## brandmilight -2.927e+01 8.741e+01
## brandmirage 4.356e+01 1.432e+02
## brandmisty 2.217e+02 3.941e+01
## brandmlife 4.214e+00 4.444e+01
## brandmonge 1.058e+01 1.432e+02
## brandmonkart 2.643e+01 1.432e+02
## brandmoshi 1.363e+02 3.364e+01
## brandmotorola -5.709e+01 1.460e+02
## brandmoulinex 9.656e+01 1.790e+01
## brandmoxom 8.844e+00 6.050e+01
## brandmsi 2.941e+02 5.009e+01
## brandmueller 5.137e-01 1.432e+02
## brandmujjo -1.251e+01 3.680e+01
## brandmuljhtidom 5.299e+00 1.854e+01
## brandnavien 6.235e+02 1.252e+02
## brandnavitel 1.213e+02 1.471e+02
## brandneo 3.836e+01 1.731e+01
## brandneoline 1.842e+02 3.792e+01
## brandneptun 5.864e+01 1.241e+02
## brandnika 9.882e+00 1.937e+01
## brandnikon 7.329e+02 3.568e+01
## brandninebot 8.981e+01 3.972e+01
## brandnintendo 6.349e+01 2.682e+01
## brandnivea 1.437e-01 2.973e+01
## brandnokia 7.058e+01 2.165e+01
## brandnokian 4.287e+01 1.432e+02
## brandnommi 2.287e+02 1.449e+02
## brandnone 1.492e+01 1.805e+01
## brandnordland 1.673e+01 3.941e+01
## brandnovatrack 1.401e+02 8.339e+01
## brandnuk 1.060e+01 1.432e+02
## brandnv-print 3.037e+02 4.594e+01
## brandnzxt 1.938e+02 7.638e+01
## brandockel 1.246e+02 1.432e+02
## brandokko -3.485e+02 2.086e+01
## brandokuma 4.279e+01 8.386e+01
## brandolimpik -1.126e+00 2.682e+01
## brandolympus 8.292e+01 1.432e+02
## brandomron 5.589e+01 2.733e+01
## brandoppo 5.782e+01 1.749e+01
## brandoptoma 5.022e+02 1.604e+02
## brandoral-b -5.197e+01 2.400e+01
## brandorgan -3.942e+02 2.089e+01
## brandorico -4.731e+01 6.140e+01
## brandosz 2.187e+01 2.431e+01
## brandowner -1.336e+00 1.432e+02
## brandozone -1.302e+01 2.433e+01
## brandpaclan 1.866e+01 7.322e+01
## brandpalisad 4.797e+00 6.050e+01
## brandpalit 7.019e+01 6.524e+01
## brandpanasonic 9.332e+01 1.830e+01
## brandpaperline 1.035e+02 1.440e+02
## brandpasabahce 9.614e-01 1.814e+01
## brandpastel 4.417e+01 2.171e+01
## brandpatriot 5.209e+01 3.040e+01
## brandpccooler 2.303e+02 1.026e+02
## brandpemco 9.569e+00 1.176e+02
## brandperilla 9.149e+00 2.189e+01
## brandpeterhof 1.279e+01 1.937e+01
## brandpetzl 3.755e+01 1.020e+02
## brandpgytech 5.556e+01 8.447e+01
## brandphantom -1.070e+02 1.847e+02
## brandphilips 7.685e+01 1.730e+01
## brandpixel 1.362e+02 1.020e+02
## brandplantronics 5.614e+01 2.798e+01
## brandplayme 3.193e+02 1.471e+02
## brandplextor 7.043e+01 1.434e+02
## brandpocketbook 9.574e+01 5.398e+01
## brandpolaris 3.965e+01 1.738e+01
## brandpolaroid -1.200e+02 7.375e+01
## brandpolimerbiht 4.480e-01 1.762e+01
## brandportcase 1.529e+01 3.836e+01
## brandpowerplant -3.247e+02 6.608e+01
## brandpowertrac 3.379e+01 5.308e+01
## brandpozis -5.650e+01 2.383e+01
## brandpresident 3.673e+02 1.434e+02
## brandprestigio -1.214e+01 3.182e+01
## brandprocab -4.406e+01 1.604e+02
## brandpromtorgservis 2.504e+01 1.432e+02
## brandproscreen 1.268e+02 1.604e+02
## brandprovence 2.716e+00 1.861e+01
## brandpyrex 2.673e+01 2.210e+01
## brandrapoo 5.732e+01 2.859e+01
## brandrastar 3.918e+01 3.482e+01
## brandrazer 1.757e+02 2.782e+01
## brandredmond 4.712e+01 1.783e+01
## brandregnum 3.259e+01 4.138e+01
## brandremington 4.527e+01 1.990e+01
## brandresanta 3.723e+01 4.801e+01
## brandresto 2.372e+01 6.052e+01
## brandrioba 9.737e-01 1.432e+02
## brandritmix 1.087e+01 3.539e+01
## brandriva 4.719e+01 1.023e+02
## brandrivacase 3.467e+01 2.538e+01
## brandroadx 4.164e+01 8.386e+01
## brandrockstar 5.745e+01 3.141e+01
## brandrondell 3.313e+01 1.778e+01
## brandrosneftjh 1.674e+00 1.432e+02
## brandrossija 1.417e+01 8.386e+01
## brandrowenta 4.239e+01 1.773e+01
## brandruggear 1.001e+02 6.134e+01
## brandsakura 2.898e+02 1.437e+02
## brandsamsonite 1.124e+02 5.638e+01
## brandsamsung 1.224e+02 1.713e+01
## brandsamura 2.458e+01 1.432e+02
## brandsanc 1.555e+02 3.344e+01
## brandsaramonic 4.465e+00 5.820e+01
## brandsatechi -2.271e+02 5.333e+01
## brandsavic -5.652e+00 8.258e+01
## brandsbs 1.713e+01 2.552e+01
## brandscarlett 2.624e+01 1.761e+01
## brandschwiizer 1.810e+01 1.432e+02
## brandscreentec -2.476e+00 5.308e+01
## brandseagate 1.021e+02 8.421e+01
## brandseasonic 2.074e+02 1.449e+02
## brandselect -1.576e+00 1.432e+02
## brandsencor 5.155e+01 2.058e+01
## brandsennheiser 3.082e+02 1.433e+02
## brandshelkovica 1.118e+01 1.432e+02
## brandship 2.264e+02 2.219e+01
## brandsibrtekh -2.564e+02 1.438e+02
## brandsigma 8.776e+01 1.467e+02
## brandsimax 1.206e+01 2.013e+01
## brandsimfer 7.442e+01 2.787e+01
## brandsinger 1.180e+02 4.442e+01
## brandsjcam 1.138e+02 6.050e+01
## brandskullcandy 1.068e+02 1.433e+02
## brandsmart 9.544e+00 1.432e+02
## brandsmeg 2.268e+02 8.395e+01
## brandsmile 5.013e+02 1.432e+02
## brandsony 1.578e+02 1.764e+01
## brandsparta 5.137e-01 1.432e+02
## brandspiegelau 2.048e+01 3.606e+01
## brandspigen -2.145e+00 3.679e+01
## brandsportop 4.602e+02 1.432e+02
## brandsports 7.195e+02 6.584e+01
## brandstaedtler 8.975e+01 2.019e+02
## brandstarline 1.692e+02 3.672e+01
## brandstatus -3.804e+02 6.593e+01
## brandstaub 1.756e+02 6.597e+01
## brandsteelseries 1.503e+02 3.202e+01
## brandstels 2.494e+02 1.432e+02
## brandsumdex 4.068e+01 2.678e+01
## brandsuperlux -7.715e+01 6.232e+01
## brandsvc -3.301e+01 1.966e+01
## brandsvetocopy 9.615e+01 3.101e+01
## brandsynology 6.476e+02 1.437e+02
## brandtacx 6.927e+02 1.528e+02
## brandtailg 4.878e+02 4.306e+01
## brandtamron 5.739e+02 7.298e+01
## brandtaoran 1.140e+02 1.528e+02
## brandtayfun -2.077e+02 2.842e+01
## brandtcl -1.106e+02 1.962e+01
## brandtechnodom 9.884e+00 1.761e+01
## brandtechnogym 4.354e+03 1.528e+02
## brandtechnomax 1.426e+02 1.434e+02
## brandtefal 7.535e+01 1.724e+01
## brandtemdan -2.464e+00 6.584e+01
## brandtenda 6.534e+00 1.432e+02
## brandthefaceshop 3.984e+00 1.908e+01
## brandthermaltake 1.653e+02 3.431e+01
## brandthermex -6.478e+01 2.475e+01
## brandthomas 1.784e+02 2.385e+01
## brandthule 8.754e+01 1.432e+02
## brandtigar 8.754e+01 1.432e+02
## brandtigernu 4.356e+01 1.432e+02
## brandtimberk 4.444e+01 1.023e+02
## brandtimson 4.001e+01 3.088e+01
## brandtion 5.829e+02 1.432e+02
## brandtoday -4.163e-01 1.432e+02
## brandtopperr -8.276e+00 2.718e+01
## brandtornado -7.332e+01 8.238e+01
## brandtoro 1.635e+00 1.863e+01
## brandtoshiba 1.624e+01 3.241e+01
## brandtoyo 7.829e+01 1.020e+02
## brandtp-link 1.166e+02 2.787e+01
## brandtramp -5.750e-01 5.769e+01
## brandtranscend 3.810e+01 1.818e+01
## brandtribe -6.992e+01 6.943e+01
## brandtrio -3.924e+02 2.422e+01
## brandtrust -1.130e+02 2.130e+01
## brandtucano 5.258e+01 3.386e+01
## brandturbo -1.573e+00 8.386e+01
## brandtvs 4.674e+01 1.796e+01
## brandtyr 2.389e+01 1.432e+02
## brandubisoft 4.472e+01 1.020e+02
## brandUnknown 3.884e+01 1.714e+01
## brandurbanears 1.593e+02 1.432e+02
## branduriage 1.102e+01 1.432e+02
## brandusams 4.671e+01 3.117e+01
## branduteki -3.903e+02 2.723e+01
## brandvarta -7.985e+01 2.363e+01
## brandventa 2.808e+02 1.020e+02
## brandviatti 4.955e+01 6.050e+01
## brandvictoria 3.353e+01 2.033e+01
## brandvikhrjh 8.266e+01 4.189e+01
## brandvirtuix 1.231e+01 1.432e+02
## brandvitehks -4.457e+00 4.815e+01
## brandvitek 2.812e+01 1.745e+01
## brandvivo 2.425e+01 2.058e+01
## brandvoin -8.488e+01 1.482e+02
## brandvortex 3.326e+01 1.022e+02
## brandwacom 7.110e+01 4.199e+01
## brandweber 1.446e+01 7.312e+01
## brandwethepeople 4.561e+02 1.521e+02
## brandwhirlpool 6.829e+02 7.320e+01
## brandwilmax 1.164e+01 2.222e+01
## brandwilson 1.092e+01 1.432e+02
## brandwintek 2.939e+01 5.308e+01
## brandwmf 3.035e+02 6.619e+01
## brandwonlex 2.649e+01 1.434e+02
## brandwxd 3.099e+01 1.023e+02
## brandx-game 4.686e+01 1.935e+01
## brandxbox 1.135e+02 2.958e+01
## brandxerox 2.807e+02 7.385e+01
## brandxiaomi 4.881e+01 1.747e+01
## brandxp-pen 7.468e+01 4.330e+01
## brandyokohama 8.292e+01 8.386e+01
## brandyonker 6.440e+01 1.432e+02
## brandyoobao 4.690e+01 2.844e+01
## brandzala 9.404e+01 7.463e+01
## brandzalman 7.063e+01 1.026e+02
## brandzeppelin 1.874e+01 1.444e+02
## brandzhiyun 2.066e+02 4.444e+01
## brandzhorka -1.186e+00 1.020e+02
## brandzowie 1.265e+02 3.975e+01
## brandzugo 2.507e+01 6.584e+01
## brandzwilling 8.430e+01 3.847e+01
## category_code0.05 -2.315e-10 3.386e+01
## category_code0.23 -2.086e-10 2.227e+01
## category_code0.35 -2.090e-10 1.946e+01
## category_code0.42 -2.083e-10 1.772e+01
## category_code0.44 -2.702e-10 1.012e+02
## category_code0.46 -2.118e-10 3.072e+01
## category_code0.51 -3.425e-10 1.012e+02
## category_code0.58 -2.080e-10 2.640e+01
## category_code0.65 -2.106e-10 4.881e+01
## category_code0.67 -2.068e-10 2.173e+01
## category_code0.69 -2.395e-10 7.206e+01
## category_code0.76 -1.669e-10 3.741e+01
## category_code0.81 -2.132e-10 1.790e+01
## category_code0.93 -1.829e-10 2.059e+01
## category_code0.97 -2.054e-10 1.708e+01
## category_code1.13 -2.181e-10 1.427e+02
## category_code1.16 -2.123e-10 1.953e+01
## category_code1.37 -1.916e-10 5.161e+01
## category_code1.39 -2.136e-10 1.012e+02
## category_code1.50 -2.066e-10 6.466e+01
## category_code1.60 -1.922e-10 3.548e+01
## category_code1.62 -2.074e-10 4.112e+01
## category_code1.74 -1.388e-10 3.852e+01
## category_code1.83 -2.866e-10 3.852e+01
## category_code10.42 -2.281e-10 1.323e+01
## category_code10.44 -1.574e-10 4.645e+01
## category_code104.17 -2.192e-10 4.112e+01
## category_code106.43 -4.878e-12 1.427e+02
## category_code11.32 -2.028e-10 7.206e+01
## category_code11.55 -1.833e-10 2.030e+01
## category_code11.57 -2.084e-10 1.484e+01
## category_code115.74 -1.290e-10 4.443e+01
## category_code118.72 -3.740e-10 1.427e+02
## category_code12.71 -1.898e-10 1.427e+02
## category_code12.73 -1.936e-10 3.741e+01
## category_code12.94 -2.114e-10 1.012e+02
## category_code120.98 -3.126e-10 6.466e+01
## category_code122.66 -2.217e-10 1.012e+02
## category_code13.66 -2.004e-10 7.206e+01
## category_code13.87 -2.101e-10 1.349e+01
## category_code13.89 -2.531e-10 2.531e+01
## category_code13.91 -1.709e-10 6.466e+01
## category_code132.41 -2.471e-10 4.112e+01
## category_code134.26 -1.566e-10 6.466e+01
## category_code138.87 -4.346e-10 1.427e+02
## category_code14.12 -2.097e-10 4.112e+01
## category_code14.14 -2.033e-10 5.161e+01
## category_code14.33 -1.436e-10 1.012e+02
## category_code148.13 -2.097e-10 5.921e+01
## category_code15.05 -2.043e-10 4.267e+01
## category_code15.28 -2.070e-10 2.845e+01
## category_code16.18 -2.130e-10 1.189e+01
## category_code16.20 -2.004e-10 2.670e+01
## category_code16.67 -1.901e-10 5.500e+01
## category_code162.04 -2.096e-10 8.293e+01
## category_code1666.64 -2.141e-10 1.427e+02
## category_code17.34 -1.702e-10 1.427e+02
## category_code17.36 -2.181e-10 1.415e+01
## category_code173.42 -2.086e-10 1.427e+02
## category_code18.03 -2.096e-10 8.293e+01
## category_code18.29 -2.294e-10 2.020e+01
## category_code18.50 -2.173e-10 5.161e+01
## category_code18.52 -1.553e-10 6.466e+01
## category_code18.98 -2.156e-10 1.859e+01
## category_code19.49 -2.123e-10 5.161e+01
## category_code19.68 -2.330e-10 3.463e+01
## category_code19.95 -2.128e-10 2.242e+01
## category_code2.29 -2.881e-10 3.127e+01
## category_code2.31 -1.666e-10 3.127e+01
## category_code2.52 -2.075e-10 4.881e+01
## category_code2.78 -2.029e-10 3.463e+01
## category_code2.99 -2.118e-10 3.021e+01
## category_code20.81 -4.373e-10 1.012e+02
## category_code20.83 -2.091e-10 2.124e+01
## category_code212.94 -1.957e-10 1.427e+02
## category_code219.88 -1.901e-10 1.427e+02
## category_code22.43 1.646e-10 1.427e+02
## category_code22.66 -2.070e-10 5.161e+01
## category_code222.22 -3.891e-10 7.206e+01
## category_code23.13 -8.518e-10 3.185e+01
## category_code23.14 -1.957e-10 1.427e+02
## category_code23.15 -2.090e-10 1.687e+01
## category_code23.33 -3.548e-10 1.012e+02
## category_code234.94 -1.998e-10 1.427e+02
## category_code24.28 -2.024e-10 8.293e+01
## category_code245.93 -1.943e-10 1.427e+02
## category_code25.44 -2.093e-10 1.012e+02
## category_code25.46 -1.852e-10 1.931e+01
## category_code26.62 -2.089e-10 3.021e+01
## category_code27.08 -2.350e-10 2.928e+01
## category_code27.29 -1.910e-10 1.427e+02
## category_code27.34 -2.086e-10 1.012e+02
## category_code27.75 -2.238e-10 8.293e+01
## category_code27.78 -2.093e-10 2.101e+01
## category_code270.67 -2.093e-10 1.427e+02
## category_code28.94 -2.110e-10 2.242e+01
## category_code29.63 -1.815e-10 1.214e+01
## category_code3.45 -2.001e-10 2.583e+01
## category_code3.47 -2.057e-10 1.535e+01
## category_code3.89 -1.762e-10 1.405e+01
## category_code30.09 -2.052e-10 1.587e+01
## category_code300.90 -1.982e-10 1.427e+02
## category_code31.23 -2.101e-10 2.186e+01
## category_code314.79 -2.031e-10 1.427e+02
## category_code32.38 -2.022e-10 1.012e+02
## category_code33.33 -2.110e-10 1.427e+02
## category_code33.54 -2.204e-10 3.247e+01
## category_code34.70 -2.281e-10 5.921e+01
## category_code34.72 -2.185e-10 1.994e+01
## category_code35.39 -2.119e-10 5.500e+01
## category_code37.04 -1.964e-10 1.772e+01
## category_code370.35 -2.917e-10 1.427e+02
## category_code38.17 -2.259e-10 4.645e+01
## category_code39.33 -2.066e-10 8.293e+01
## category_code39.56 -2.737e-10 1.427e+02
## category_code4.14 -2.248e-10 2.885e+01
## category_code4.38 -2.070e-10 1.427e+02
## category_code4.61 -2.708e-10 5.161e+01
## category_code4.63 -2.041e-10 1.258e+01
## category_code40.49 -1.608e-10 6.466e+01
## category_code41.64 -2.168e-10 7.206e+01
## category_code41.67 -1.979e-10 2.670e+01
## category_code43.96 -2.069e-10 1.427e+02
## category_code46.27 -2.626e-10 7.206e+01
## category_code46.30 -1.910e-10 4.645e+01
## category_code472.02 -3.898e-10 1.427e+02
## category_code486.09 -2.090e-10 1.427e+02
## category_code5.30 -2.247e-10 1.012e+02
## category_code5.53 -2.758e-10 1.012e+02
## category_code5.56 -2.217e-10 1.427e+02
## category_code5.79 -2.151e-10 1.805e+01
## category_code50.90 -2.176e-10 6.466e+01
## category_code50.93 -2.205e-10 2.186e+01
## category_code53.22 -4.625e-10 1.427e+02
## category_code53.24 -2.574e-10 5.161e+01
## category_code532.38 -2.080e-10 1.427e+02
## category_code55.53 -2.139e-10 1.427e+02
## category_code555.30 -9.195e-11 1.427e+02
## category_code555.53 -2.247e-10 1.427e+02
## category_code57.87 -2.151e-10 1.207e+01
## category_code578.68 -1.771e-10 1.427e+02
## category_code6.02 -2.111e-10 2.397e+01
## category_code6.71 -2.143e-10 5.161e+01
## category_code6.92 -2.095e-10 2.227e+01
## category_code6.94 -1.522e-10 1.254e+01
## category_code64.79 -2.544e-10 8.293e+01
## category_code671.27 -2.257e-10 1.427e+02
## category_code69.42 -1.948e-10 5.161e+01
## category_code69.44 -2.101e-10 1.387e+01
## category_code7.85 -2.202e-10 1.427e+02
## category_code729.14 -2.309e-10 1.427e+02
## category_code74.07 -2.129e-10 8.293e+01
## category_code773.13 -1.315e-10 1.427e+02
## category_code78.68 -2.258e-10 5.500e+01
## category_code8.08 -2.055e-10 1.012e+02
## category_code8.10 -2.047e-10 1.207e+01
## category_code8.33 -2.276e-10 4.112e+01
## category_code8.54 -1.333e-10 8.293e+01
## category_code8.59 -3.399e-10 5.921e+01
## category_code8.77 -2.151e-10 1.427e+02
## category_code80.76 -2.277e-10 8.293e+01
## category_code83.31 -2.132e-10 1.427e+02
## category_code83.33 -2.138e-10 3.975e+01
## category_code87.94 -2.316e-10 1.427e+02
## category_code9.14 -2.155e-10 5.921e+01
## category_code9.24 -2.139e-10 2.417e+01
## category_code9.26 -2.164e-10 1.211e+01
## category_code92.57 -2.721e-10 1.012e+02
## category_code92.59 -2.186e-10 1.363e+01
## category_code97.20 -1.999e-10 8.293e+01
## category_codeaccessories.bag 1.940e+01 1.400e+01
## category_codeaccessories.umbrella 7.752e+01 1.012e+02
## category_codeapparel.glove 2.968e+02 1.696e+01
## category_codeapparel.shirt 5.758e+01 2.727e+01
## category_codeapparel.sock NA NA
## category_codeapparel.trousers 5.345e+01 3.700e+01
## category_codeapparel.tshirt 4.675e+01 2.435e+01
## category_codeappliances.environment.air_conditioner 2.520e+02 1.273e+01
## category_codeappliances.environment.air_heater 5.803e+01 1.438e+01
## category_codeappliances.environment.climate -9.435e+01 5.321e+01
## category_codeappliances.environment.fan 4.572e+01 1.366e+01
## category_codeappliances.environment.vacuum 6.162e+01 1.217e+01
## category_codeappliances.environment.water_heater 2.109e+02 1.902e+01
## category_codeappliances.iron 6.051e+01 1.235e+01
## category_codeappliances.ironing_board 6.850e+01 1.698e+01
## category_codeappliances.kitchen.blender 3.407e+01 1.247e+01
## category_codeappliances.kitchen.coffee_grinder -3.712e+01 1.657e+01
## category_codeappliances.kitchen.coffee_machine 1.943e+01 1.639e+01
## category_codeappliances.kitchen.dishwasher 3.564e+02 1.417e+01
## category_codeappliances.kitchen.fryer 6.953e+01 3.651e+01
## category_codeappliances.kitchen.grill 1.543e+02 1.565e+01
## category_codeappliances.kitchen.hood 1.104e+02 1.258e+01
## category_codeappliances.kitchen.juicer 8.284e+01 1.517e+01
## category_codeappliances.kitchen.kettle 2.376e+01 1.212e+01
## category_codeappliances.kitchen.meat_grinder 6.253e+01 1.300e+01
## category_codeappliances.kitchen.microwave 6.954e+01 1.252e+01
## category_codeappliances.kitchen.mixer 1.006e+01 1.333e+01
## category_codeappliances.kitchen.oven 2.792e+02 1.331e+01
## category_codeappliances.kitchen.refrigerators 4.345e+02 1.227e+01
## category_codeappliances.kitchen.steam_cooker 5.470e+01 2.047e+01
## category_codeappliances.kitchen.toster -1.295e+01 1.499e+01
## category_codeappliances.kitchen.washer 2.518e+02 1.227e+01
## category_codeappliances.personal.hair_cutter 1.958e+01 1.367e+01
## category_codeappliances.personal.massager 3.709e+01 1.519e+01
## category_codeappliances.personal.scales 2.659e+01 1.248e+01
## category_codeappliances.sewing_machine 8.435e+01 2.516e+01
## category_codeappliances.steam_cleaner 3.974e+01 6.619e+01
## category_codeauto.accessories.alarm NA NA
## category_codeauto.accessories.anti_freeze 3.476e+01 1.176e+02
## category_codeauto.accessories.compressor 1.468e+02 3.973e+01
## category_codeauto.accessories.player NA NA
## category_codeauto.accessories.radar -4.329e+01 5.453e+01
## category_codeauto.accessories.videoregister -2.673e+00 3.530e+01
## category_codecomputers.components.cdrw 3.081e+01 3.382e+01
## category_codecomputers.components.cooler -1.795e+02 1.595e+01
## category_codecomputers.components.cpu 1.616e+02 1.517e+01
## category_codecomputers.components.hdd 2.694e+01 1.403e+01
## category_codecomputers.components.memory 4.646e+01 2.190e+01
## category_codecomputers.components.motherboard -1.266e+01 2.867e+01
## category_codecomputers.components.power_supply 9.297e+01 2.470e+01
## category_codecomputers.components.sound_card 9.215e+01 1.544e+02
## category_codecomputers.components.videocards 2.216e+02 2.709e+01
## category_codecomputers.desktop 2.360e+02 1.881e+01
## category_codecomputers.ebooks 1.428e+02 5.712e+01
## category_codecomputers.gaming 2.797e+01 3.460e+01
## category_codecomputers.network.router -4.634e+01 2.465e+01
## category_codecomputers.notebook 3.967e+02 1.303e+01
## category_codecomputers.peripherals.camera 4.518e+01 1.868e+01
## category_codecomputers.peripherals.joystick -1.907e+01 1.673e+01
## category_codecomputers.peripherals.keyboard 1.618e+01 1.382e+01
## category_codecomputers.peripherals.monitor 9.383e+01 1.478e+01
## category_codecomputers.peripherals.mouse -2.446e+00 1.268e+01
## category_codecomputers.peripherals.printer -8.406e+01 1.564e+01
## category_codecomputers.peripherals.scanner -1.608e+02 4.988e+01
## category_codeconstruction.components.faucet 1.441e+01 9.223e+01
## category_codeconstruction.tools.drill -5.565e+00 3.857e+01
## category_codeconstruction.tools.generator 1.731e+02 9.132e+01
## category_codeconstruction.tools.heater 5.598e+01 1.140e+02
## category_codeconstruction.tools.light 1.517e+01 1.746e+02
## category_codeconstruction.tools.pump 1.515e+02 9.137e+01
## category_codeconstruction.tools.saw 4.615e+01 6.499e+01
## category_codeconstruction.tools.screw -5.423e+01 1.347e+01
## category_codeconstruction.tools.welding 6.674e+01 7.464e+01
## category_codecountry_yard.cultivator 2.010e+02 9.132e+01
## category_codecountry_yard.lawn_mower 5.911e+01 5.960e+01
## category_codecountry_yard.watering -1.120e+02 1.443e+02
## category_codecountry_yard.weather_station 2.522e+01 6.088e+01
## category_codeelectronics.audio.acoustic 2.167e+02 1.925e+01
## category_codeelectronics.audio.dictaphone -4.954e+01 1.428e+02
## category_codeelectronics.audio.headphone -5.644e+01 1.221e+01
## category_codeelectronics.audio.microphone 1.045e+02 1.850e+01
## category_codeelectronics.audio.subwoofer 2.454e+02 1.172e+02
## category_codeelectronics.camera.photo 2.983e+02 3.361e+01
## category_codeelectronics.camera.video 3.280e+02 7.233e+01
## category_codeelectronics.clocks 5.168e+01 1.339e+01
## category_codeelectronics.smartphone 1.807e+02 1.198e+01
## category_codeelectronics.tablet 1.324e+02 1.295e+01
## category_codeelectronics.telephone -5.949e+00 1.553e+01
## category_codeelectronics.video.projector 9.213e+01 7.306e+01
## category_codeelectronics.video.tv 4.031e+02 1.227e+01
## category_codefurniture.bathroom.bath 8.143e+01 2.733e+01
## category_codefurniture.bathroom.toilet -7.426e+01 7.560e+01
## category_codefurniture.bedroom.bed 3.620e+01 1.958e+01
## category_codefurniture.bedroom.blanket 2.677e+02 2.867e+01
## category_codefurniture.bedroom.pillow 1.582e+00 1.940e+01
## category_codefurniture.kitchen.chair 3.801e+01 1.338e+01
## category_codefurniture.kitchen.table 2.752e+01 1.221e+01
## category_codefurniture.living_room.cabinet 8.279e+00 1.290e+01
## category_codefurniture.living_room.chair 1.466e+02 3.048e+01
## category_codefurniture.living_room.shelving 2.670e+01 1.358e+01
## category_codefurniture.living_room.sofa 8.304e+01 3.078e+01
## category_codekids.bottles 2.913e+01 1.464e+02
## category_codekids.carriage 2.154e+01 2.015e+02
## category_codekids.dolls 7.823e+01 4.172e+01
## category_codekids.skates 4.282e+02 2.498e+01
## category_codekids.swing 1.384e+02 1.464e+02
## category_codekids.toys 3.326e+01 3.171e+01
## category_codemedicine.tools.tonometer 2.561e+01 1.672e+01
## category_codesport.bicycle 6.882e+01 5.253e+01
## category_codesport.tennis -4.544e+01 4.459e+01
## category_codesport.trainer 1.794e+02 5.439e+01
## category_codestationery.battery 1.232e+02 1.867e+01
## category_codestationery.cartrige -2.498e+02 1.654e+01
## category_codestationery.paper -5.315e+01 1.891e+01
## category_codestationery.stapler 4.249e+00 4.682e+01
## category_codeUnknown 4.154e+01 1.182e+01
## t value Pr(>|t|)
## (Intercept) -1.876 0.060687 .
## brandaccesstyle 0.497 0.619145
## brandaction 0.159 0.873898
## brandadidas 0.386 0.699261
## brandadvantek 0.430 0.667544
## brandaeg 8.865 < 2e-16 ***
## brandaerocool 1.130 0.258436
## brandaimoto 1.497 0.134517
## brandairline 0.308 0.758102
## brandakku 0.901 0.367395
## brandaksion 1.383 0.166523
## brandakvafor 0.379 0.704899
## brandalex 0.472 0.636759
## brandalfa 0.077 0.938658
## brandalienware 25.256 < 2e-16 ***
## brandalmatv 0.200 0.841296
## brandaltair 0.239 0.810732
## brandaltel 4.492 7.06e-06 ***
## brandaltex 4.654 3.26e-06 ***
## brandamazon 1.202 0.229304
## brandamd 2.763 0.005729 **
## brandamigami -0.011 0.991164
## brandanymode 1.332 0.182739
## brandaoc -6.958 3.46e-12 ***
## brandaoki 1.577 0.114848
## brandapc -4.699 2.61e-06 ***
## brandapollo 0.325 0.745257
## brandapple 30.829 < 2e-16 ***
## brandaqua -3.006 0.002645 **
## brandaquadent 1.062 0.288366
## brandaquael 0.248 0.804460
## brandaquapick 1.823 0.068348 .
## brandarena 0.125 0.900698
## brandariston -1.826 0.067818 .
## brandarktika 0.845 0.397848
## brandarnica 0.454 0.650054
## brandart.fit 0.447 0.655085
## brandartberry -0.004 0.997180
## brandartel 2.823 0.004753 **
## brandasrock 2.597 0.009412 **
## brandastonish 0.165 0.868688
## brandasus 10.887 < 2e-16 ***
## brandatlant -2.759 0.005795 **
## brandatmor -2.744 0.006074 **
## brandatom 0.143 0.886620
## brandaudac -0.295 0.767903
## brandaura -0.025 0.980084
## brandausini 0.320 0.748601
## brandauthor 1.262 0.206907
## brandava 0.372 0.709859
## brandavermedia 1.773 0.076278 .
## brandavrora -0.042 0.966702
## brandawax -7.323 2.44e-13 ***
## brandawei 2.951 0.003165 **
## brandbaboo 0.143 0.886620
## brandbabyliss 2.303 0.021261 *
## brandbaltextile 0.971 0.331726
## brandbarbie -0.200 0.841511
## brandbardahl 0.149 0.881550
## brandbarjher 0.207 0.836016
## brandbarkan 1.797 0.072395 .
## brandbeats 4.901 9.54e-07 ***
## brandbeeline 4.056 4.99e-05 ***
## brandbeko 2.102 0.035588 *
## brandbelcando -0.189 0.849806
## brandbelita -0.147 0.882885
## brandbelitam 0.003 0.997918
## brandbellissima 1.703 0.088538 .
## brandbequiet 3.814 0.000137 ***
## brandberghoff 1.197 0.231474
## brandberkley 0.018 0.985553
## brandberry 0.239 0.810732
## brandbestway 0.423 0.672098
## brandbeurer 1.880 0.060079 .
## brandbfgoodrich 0.374 0.708657
## brandbioderma 0.151 0.880118
## brandbiol 0.133 0.894226
## brandbiolane 0.336 0.736550
## brandbiostar 1.155 0.248056
## brandbirjusa -6.083 1.18e-09 ***
## brandbisfree 0.007 0.994576
## brandblackstar 1.750 0.080131 .
## brandblehk 0.294 0.768846
## brandbloody 4.155 3.25e-05 ***
## brandbona 2.493 0.012682 *
## brandborasco -0.053 0.957466
## brandbork 14.174 < 2e-16 ***
## brandborner 1.107 0.268509
## brandbosch 11.668 < 2e-16 ***
## brandbose 7.008 2.43e-12 ***
## brandbradex 0.299 0.764815
## brandbrateck 0.951 0.341815
## brandbraun 4.947 7.54e-07 ***
## brandbrelil 0.057 0.954607
## brandbridgestone 1.128 0.259425
## brandbrother 8.633 < 2e-16 ***
## brandbruder 0.372 0.709852
## brandbuebchen 0.013 0.989285
## brandbuff 0.146 0.884084
## brandbushido 0.259 0.795337
## brandbyintek 1.688 0.091359 .
## brandbykski 0.521 0.602246
## brandcablexpert 0.061 0.951392
## brandcalgon 0.034 0.972635
## brandcamelion -3.971 7.15e-05 ***
## brandcannondale 6.649 2.96e-11 ***
## brandcanon 15.580 < 2e-16 ***
## brandcanyon -3.237 0.001209 **
## brandcasada 4.043 5.28e-05 ***
## brandcasio -0.585 0.558328
## brandcaso 3.393 0.000692 ***
## brandcaspio 1.594 0.110971
## brandcatrice -0.004 0.996456
## brandcdc 0.154 0.877904
## brandcelebrat 0.667 0.505050
## brandchicco 0.441 0.659121
## brandchina -16.511 < 2e-16 ***
## brandcilek 26.361 < 2e-16 ***
## brandcliny -1.771 0.076535 .
## brandcollecta 0.005 0.995981
## brandcolorful 0.921 0.356823
## brandcompliment -0.318 0.750260
## brandcontinent -3.052 0.002274 **
## brandcontinental 1.451 0.146878
## brandcoolinar 0.787 0.431281
## brandcort 1.587 0.112452
## brandcougar 1.321 0.186636
## brandcremesso 0.324 0.746175
## brandcullmann -2.950 0.003174 **
## brandcyberpower -2.456 0.014031 *
## brandd-color 0.094 0.925059
## brandd-link 2.513 0.011983 *
## branddaewoo 1.467 0.142480
## branddaikin 2.550 0.010768 *
## branddaiwa 0.270 0.787007
## branddam -0.014 0.988658
## branddarina 2.243 0.024902 *
## branddc-girls -0.504 0.614208
## branddecoroom 0.403 0.686977
## branddeepcool -5.631 1.79e-08 ***
## branddeerma 0.359 0.719548
## branddefender 0.270 0.787435
## branddell 6.295 3.09e-10 ***
## branddelonghi 11.980 < 2e-16 ***
## branddelux 3.269 0.001081 **
## branddeluxe 2.001 0.045428 *
## branddemidovskiy 0.141 0.887893
## branddepileve 0.013 0.989619
## branddermal -0.021 0.983100
## branddermoviva 0.022 0.982121
## branddesignskin -0.013 0.989963
## branddifferent -6.157 7.43e-10 ***
## branddji 18.499 < 2e-16 ***
## branddogland 2.945 0.003228 **
## branddomini 1.192 0.233224
## branddougez 0.053 0.957617
## branddr.beckmann -0.008 0.993405
## brandduracell -3.049 0.002299 **
## branddxracer 4.599 4.25e-06 ***
## branddyson 22.718 < 2e-16 ***
## brande.gov 0.402 0.687576
## brandecocool 1.948 0.051432 .
## brandecologystone 0.928 0.353347
## brandedifier 2.920 0.003499 **
## brandeglo 0.332 0.739923
## brandegoiste -0.006 0.995457
## brandehlektrostandart 0.159 0.873898
## brandehra -3.710 0.000208 ***
## brandelari 1.296 0.195052
## brandelectrolux 15.656 < 2e-16 ***
## brandemsa 0.905 0.365542
## brandenergea -0.221 0.825118
## brandepson 14.347 < 2e-16 ***
## brandergolux -5.341 9.25e-08 ***
## brandeset -10.435 < 2e-16 ***
## brandetalon -0.455 0.648897
## brandeuroprint 4.567 4.96e-06 ***
## brandeverlast 0.401 0.688349
## brandeyfel -0.890 0.373427
## brandezviz 0.694 0.487686
## brandfender 2.888 0.003876 **
## brandfiltero -0.125 0.900673
## brandfissman 0.879 0.379369
## brandfitokosmetik -0.012 0.990607
## brandfixsen 0.090 0.928110
## brandfizan 0.092 0.926335
## brandflama 4.175 2.98e-05 ***
## brandfly -1.306 0.191665
## brandforward 1.355 0.175550
## brandfossil 2.555 0.010616 *
## brandfranke 1.310 0.190030
## brandfujifilm 9.853 < 2e-16 ***
## brandgalaxy 0.320 0.748809
## brandgamdias 1.175 0.239893
## brandgamemax 1.535 0.124843
## brandgarmin 25.015 < 2e-16 ***
## brandgefest 10.742 < 2e-16 ***
## brandgembird 0.228 0.819549
## brandgenau 2.566 0.010291 *
## brandgeneris 0.708 0.478865
## brandgenius 1.625 0.104076
## brandgerat 0.231 0.817009
## brandgewa 0.603 0.546524
## brandgeyzer 0.035 0.971841
## brandgezatone 1.843 0.065355 .
## brandggg 0.617 0.537492
## brandgiant 4.563 5.06e-06 ***
## brandgigabyte 6.338 2.33e-10 ***
## brandgillette 0.077 0.938258
## brandgiottos -2.058 0.039602 *
## brandglasslock 0.295 0.768226
## brandglassware 0.669 0.503224
## brandglobal 3.344 0.000827 ***
## brandgo-sport 5.435 5.49e-08 ***
## brandgolf -0.011 0.991387
## brandgoodride 2.473 0.013411 *
## brandgoodyear -0.875 0.381483
## brandgopro 5.337 9.48e-08 ***
## brandgorenje 3.205 0.001352 **
## brandgrans 1.522 0.128092
## brandgreenway 3.475 0.000512 ***
## brandgrillver 0.433 0.664710
## brandgrohe 9.059 < 2e-16 ***
## brandgtec 0.147 0.882904
## brandgutrend 6.945 3.80e-12 ***
## brandhabilead 2.185 0.028905 *
## brandhansa 10.733 < 2e-16 ***
## brandhaushalt 0.332 0.739803
## brandhenkel 0.077 0.938864
## brandherschel 0.849 0.395636
## brandhintek 0.442 0.658270
## brandhitachi 17.512 < 2e-16 ***
## brandhms 0.898 0.369345
## brandhoco -0.044 0.964515
## brandhonor -0.514 0.607078
## brandhotpoint-ariston 4.833 1.35e-06 ***
## brandhp 14.719 < 2e-16 ***
## brandhtc 6.405 1.51e-10 ***
## brandhuawei 4.164 3.14e-05 ***
## brandhuion -0.123 0.902067
## brandhuntkey -10.718 < 2e-16 ***
## brandhurom 2.507 0.012162 *
## brandhuter 3.182 0.001466 **
## brandhygge 4.409 1.04e-05 ***
## brandhyperx 6.317 2.68e-10 ***
## brandhyundai -0.367 0.713885
## brandid-cooling 3.489 0.000485 ***
## brandideal 0.043 0.965942
## brandikins -0.003 0.997681
## brandimetec 0.953 0.340431
## brandincase 4.564 5.02e-06 ***
## brandindesit -1.794 0.072806 .
## brandinhouse 0.790 0.429611
## brandinkax 0.676 0.498840
## brandinoi 0.489 0.624819
## brandinsight -0.699 0.484732
## brandinspector 2.751 0.005940 **
## brandintel 2.985 0.002835 **
## brandintex 3.935 8.32e-05 ***
## brandionkini 0.136 0.892026
## brandipower -17.495 < 2e-16 ***
## brandiqos 1.013 0.311219
## brandirbis -3.276 0.001054 **
## brandivi -15.563 < 2e-16 ***
## brandivolia 0.490 0.624190
## brandiwalk 0.011 0.990987
## brandjabra 1.493 0.135323
## brandjaguar 3.158 0.001587 **
## brandjandeks 4.630 3.66e-06 ***
## brandjanome 4.926 8.42e-07 ***
## brandjbl 6.552 5.70e-11 ***
## brandjetair -0.234 0.815273
## brandjetpik 0.530 0.595843
## brandjoby -4.643 3.44e-06 ***
## brandjoerex -4.233 2.31e-05 ***
## brandjonsbo 1.299 0.194052
## brandjvc 0.896 0.370000
## brandkaabo 3.494 0.000475 ***
## brandkama 0.377 0.705888
## brandkarcher 5.696 1.23e-08 ***
## brandkaspersky -8.159 3.40e-16 ***
## brandkenko -2.767 0.005650 **
## brandkenwood 9.209 < 2e-16 ***
## brandkicx -0.744 0.457077
## brandking 2.933 0.003355 **
## brandkingston 2.017 0.043713 *
## brandkitchenaid 13.983 < 2e-16 ***
## brandkivi -3.669 0.000244 ***
## brandkmk 0.019 0.985109
## brandkomax 0.100 0.920640
## brandkosadaka -0.009 0.992557
## brandkramet -9.772 < 2e-16 ***
## brandkrups 15.212 < 2e-16 ***
## brandkumano -2.492 0.012694 *
## brandkumho 0.395 0.693084
## brandkurzweil 5.123 3.01e-07 ***
## brandkyocera 2.824 0.004742 **
## brandlamart 0.039 0.969103
## brandlaurastar 12.011 < 2e-16 ***
## brandlavazza 0.412 0.680657
## brandlego 1.380 0.167735
## brandlegrand -9.103 < 2e-16 ***
## brandlenovo 10.921 < 2e-16 ***
## brandlenspen 0.109 0.912845
## brandlg 11.050 < 2e-16 ***
## brandlihom 3.463 0.000535 ***
## brandlion -1.355 0.175296
## brandloewe 29.768 < 2e-16 ***
## brandlogitech 3.438 0.000586 ***
## brandlol -0.218 0.827179
## brandlori -0.008 0.993224
## brandlotte 4.478 7.55e-06 ***
## brandlowepro 0.474 0.635615
## brandluch -0.012 0.990329
## brandlumax 0.374 0.708618
## brandluminarc 2.392 0.016760 *
## brandluxell 0.138 0.890015
## brandmaestro 1.027 0.304301
## brandmakita 4.485 7.31e-06 ***
## brandmanfrotto -1.275 0.202292
## brandmarcato 2.438 0.014773 *
## brandmarcel 0.004 0.997139
## brandmarley 1.134 0.256678
## brandmarshall 3.724 0.000196 ***
## brandmart 6.722 1.80e-11 ***
## brandmarvel 0.385 0.700260
## brandmatrix 0.102 0.918912
## brandmattel 1.587 0.112536
## brandmaxwell 1.219 0.222699
## brandmcdavid 0.097 0.922805
## brandmedisana 0.536 0.591657
## brandmegogo -17.401 < 2e-16 ***
## brandmercusys 2.901 0.003718 **
## brandmetabo 0.532 0.595071
## brandmetalions 0.164 0.869839
## brandmichelin 0.862 0.388883
## brandmicro -1.089 0.275992
## brandmicrolab -10.677 < 2e-16 ***
## brandmicrosoft -8.432 < 2e-16 ***
## brandmidea -5.047 4.50e-07 ***
## brandmilight -0.335 0.737747
## brandmirage 0.304 0.761031
## brandmisty 5.625 1.86e-08 ***
## brandmlife 0.095 0.924456
## brandmonge 0.074 0.941100
## brandmonkart 0.185 0.853590
## brandmoshi 4.052 5.09e-05 ***
## brandmotorola -0.391 0.695724
## brandmoulinex 5.395 6.86e-08 ***
## brandmoxom 0.146 0.883786
## brandmsi 5.871 4.34e-09 ***
## brandmueller 0.004 0.997139
## brandmujjo -0.340 0.733876
## brandmuljhtidom 0.286 0.774947
## brandnavien 4.978 6.43e-07 ***
## brandnavitel 0.825 0.409270
## brandneo 2.216 0.026671 *
## brandneoline 4.856 1.20e-06 ***
## brandneptun 0.473 0.636558
## brandnika 0.510 0.609979
## brandnikon 20.542 < 2e-16 ***
## brandninebot 2.261 0.023769 *
## brandnintendo 2.367 0.017927 *
## brandnivea 0.005 0.996143
## brandnokia 3.261 0.001111 **
## brandnokian 0.299 0.764703
## brandnommi 1.578 0.114460
## brandnone 0.827 0.408476
## brandnordland 0.424 0.671221
## brandnovatrack 1.680 0.092895 .
## brandnuk 0.074 0.940989
## brandnv-print 6.611 3.83e-11 ***
## brandnzxt 2.537 0.011167 *
## brandockel 0.870 0.384440
## brandokko -16.703 < 2e-16 ***
## brandokuma 0.510 0.609830
## brandolimpik -0.042 0.966518
## brandolympus 0.579 0.562652
## brandomron 2.045 0.040867 *
## brandoppo 3.306 0.000948 ***
## brandoptoma 3.132 0.001738 **
## brandoral-b -2.165 0.030368 *
## brandorgan -18.869 < 2e-16 ***
## brandorico -0.771 0.440954
## brandosz 0.900 0.368206
## brandowner -0.009 0.992557
## brandozone -0.535 0.592504
## brandpaclan 0.255 0.798829
## brandpalisad 0.079 0.936804
## brandpalit 1.076 0.282009
## brandpanasonic 5.100 3.40e-07 ***
## brandpaperline 0.719 0.472224
## brandpasabahce 0.053 0.957720
## brandpastel 2.035 0.041862 *
## brandpatriot 1.714 0.086613 .
## brandpccooler 2.245 0.024773 *
## brandpemco 0.081 0.935142
## brandperilla 0.418 0.676056
## brandpeterhof 0.660 0.509197
## brandpetzl 0.368 0.712775
## brandpgytech 0.658 0.510650
## brandphantom -0.580 0.562211
## brandphilips 4.442 8.93e-06 ***
## brandpixel 1.335 0.181889
## brandplantronics 2.006 0.044818 *
## brandplayme 2.171 0.029933 *
## brandplextor 0.491 0.623435
## brandpocketbook 1.774 0.076125 .
## brandpolaris 2.281 0.022555 *
## brandpolaroid -1.627 0.103669
## brandpolimerbiht 0.025 0.979721
## brandportcase 0.399 0.690210
## brandpowerplant -4.914 8.93e-07 ***
## brandpowertrac 0.636 0.524452
## brandpozis -2.371 0.017736 *
## brandpresident 2.562 0.010395 *
## brandprestigio -0.382 0.702716
## brandprocab -0.275 0.783518
## brandpromtorgservis 0.175 0.861209
## brandproscreen 0.790 0.429245
## brandprovence 0.146 0.883990
## brandpyrex 1.209 0.226571
## brandrapoo 2.005 0.044969 *
## brandrastar 1.125 0.260475
## brandrazer 6.315 2.71e-10 ***
## brandredmond 2.643 0.008224 **
## brandregnum 0.788 0.430958
## brandremington 2.275 0.022905 *
## brandresanta 0.775 0.438063
## brandresto 0.392 0.695120
## brandrioba 0.007 0.994576
## brandritmix 0.307 0.758788
## brandriva 0.461 0.644499
## brandrivacase 1.366 0.171880
## brandroadx 0.497 0.619498
## brandrockstar 1.829 0.067373 .
## brandrondell 1.864 0.062377 .
## brandrosneftjh 0.012 0.990677
## brandrossija 0.169 0.865811
## brandrowenta 2.391 0.016796 *
## brandruggear 1.632 0.102783
## brandsakura 2.016 0.043755 *
## brandsamsonite 1.993 0.046294 *
## brandsamsung 7.147 8.90e-13 ***
## brandsamura 0.172 0.863733
## brandsanc 4.648 3.35e-06 ***
## brandsaramonic 0.077 0.938849
## brandsatechi -4.259 2.05e-05 ***
## brandsavic -0.068 0.945435
## brandsbs 0.671 0.502110
## brandscarlett 1.490 0.136192
## brandschwiizer 0.126 0.899426
## brandscreentec -0.047 0.962792
## brandseagate 1.213 0.225118
## brandseasonic 1.432 0.152246
## brandselect -0.011 0.991220
## brandsencor 2.504 0.012272 *
## brandsennheiser 2.151 0.031467 *
## brandshelkovica 0.078 0.937768
## brandship 10.202 < 2e-16 ***
## brandsibrtekh -1.783 0.074556 .
## brandsigma 0.598 0.549583
## brandsimax 0.599 0.549136
## brandsimfer 2.670 0.007580 **
## brandsinger 2.656 0.007898 **
## brandsjcam 1.881 0.060025 .
## brandskullcandy 0.746 0.455927
## brandsmart 0.067 0.946879
## brandsmeg 2.701 0.006906 **
## brandsmile 3.499 0.000467 ***
## brandsony 8.946 < 2e-16 ***
## brandsparta 0.004 0.997139
## brandspiegelau 0.568 0.570145
## brandspigen -0.058 0.953516
## brandsportop 3.213 0.001314 **
## brandsports 10.928 < 2e-16 ***
## brandstaedtler 0.444 0.656685
## brandstarline 4.608 4.07e-06 ***
## brandstatus -5.769 8.00e-09 ***
## brandstaub 2.662 0.007778 **
## brandsteelseries 4.695 2.67e-06 ***
## brandstels 1.741 0.081722 .
## brandsumdex 1.519 0.128811
## brandsuperlux -1.238 0.215752
## brandsvc -1.679 0.093080 .
## brandsvetocopy 3.100 0.001934 **
## brandsynology 4.506 6.61e-06 ***
## brandtacx 4.534 5.79e-06 ***
## brandtailg 11.328 < 2e-16 ***
## brandtamron 7.864 3.76e-15 ***
## brandtaoran 0.746 0.455520
## brandtayfun -7.308 2.72e-13 ***
## brandtcl -5.639 1.71e-08 ***
## brandtechnodom 0.561 0.574559
## brandtechnogym 28.497 < 2e-16 ***
## brandtechnomax 0.995 0.319980
## brandtefal 4.372 1.23e-05 ***
## brandtemdan -0.037 0.970143
## brandtenda 0.046 0.963619
## brandthefaceshop 0.209 0.834644
## brandthermaltake 4.817 1.46e-06 ***
## brandthermex -2.618 0.008859 **
## brandthomas 7.479 7.54e-14 ***
## brandthule 0.611 0.541093
## brandtigar 0.611 0.541093
## brandtigernu 0.304 0.761031
## brandtimberk 0.434 0.664060
## brandtimson 1.296 0.195063
## brandtion 4.070 4.71e-05 ***
## brandtoday -0.003 0.997681
## brandtopperr -0.304 0.760748
## brandtornado -0.890 0.373479
## brandtoro 0.088 0.930081
## brandtoshiba 0.501 0.616257
## brandtoyo 0.768 0.442754
## brandtp-link 4.185 2.85e-05 ***
## brandtramp -0.010 0.992047
## brandtranscend 2.095 0.036161 *
## brandtribe -1.007 0.313911
## brandtrio -16.206 < 2e-16 ***
## brandtrust -5.304 1.13e-07 ***
## brandtucano 1.553 0.120488
## brandturbo -0.019 0.985035
## brandtvs 2.602 0.009267 **
## brandtyr 0.167 0.867522
## brandubisoft 0.438 0.661041
## brandUnknown 2.266 0.023465 *
## brandurbanears 1.112 0.266082
## branduriage 0.077 0.938656
## brandusams 1.499 0.133927
## branduteki -14.336 < 2e-16 ***
## brandvarta -3.379 0.000727 ***
## brandventa 2.753 0.005899 **
## brandviatti 0.819 0.412843
## brandvictoria 1.649 0.099073 .
## brandvikhrjh 1.973 0.048474 *
## brandvirtuix 0.086 0.931495
## brandvitehks -0.093 0.926240
## brandvitek 1.612 0.106995
## brandvivo 1.179 0.238554
## brandvoin -0.573 0.566774
## brandvortex 0.326 0.744793
## brandwacom 1.693 0.090466 .
## brandweber 0.198 0.843245
## brandwethepeople 2.998 0.002714 **
## brandwhirlpool 9.329 < 2e-16 ***
## brandwilmax 0.524 0.600265
## brandwilson 0.076 0.939212
## brandwintek 0.554 0.579816
## brandwmf 4.585 4.55e-06 ***
## brandwonlex 0.185 0.853437
## brandwxd 0.303 0.761906
## brandx-game 2.421 0.015463 *
## brandxbox 3.836 0.000125 ***
## brandxerox 3.802 0.000144 ***
## brandxiaomi 2.795 0.005197 **
## brandxp-pen 1.725 0.084567 .
## brandyokohama 0.989 0.322749
## brandyonker 0.450 0.652988
## brandyoobao 1.649 0.099114 .
## brandzala 1.260 0.207637
## brandzalman 0.689 0.491092
## brandzeppelin 0.130 0.896752
## brandzhiyun 4.648 3.35e-06 ***
## brandzhorka -0.012 0.990720
## brandzowie 3.182 0.001465 **
## brandzugo 0.381 0.703328
## brandzwilling 2.191 0.028440 *
## category_code0.05 0.000 1.000000
## category_code0.23 0.000 1.000000
## category_code0.35 0.000 1.000000
## category_code0.42 0.000 1.000000
## category_code0.44 0.000 1.000000
## category_code0.46 0.000 1.000000
## category_code0.51 0.000 1.000000
## category_code0.58 0.000 1.000000
## category_code0.65 0.000 1.000000
## category_code0.67 0.000 1.000000
## category_code0.69 0.000 1.000000
## category_code0.76 0.000 1.000000
## category_code0.81 0.000 1.000000
## category_code0.93 0.000 1.000000
## category_code0.97 0.000 1.000000
## category_code1.13 0.000 1.000000
## category_code1.16 0.000 1.000000
## category_code1.37 0.000 1.000000
## category_code1.39 0.000 1.000000
## category_code1.50 0.000 1.000000
## category_code1.60 0.000 1.000000
## category_code1.62 0.000 1.000000
## category_code1.74 0.000 1.000000
## category_code1.83 0.000 1.000000
## category_code10.42 0.000 1.000000
## category_code10.44 0.000 1.000000
## category_code104.17 0.000 1.000000
## category_code106.43 0.000 1.000000
## category_code11.32 0.000 1.000000
## category_code11.55 0.000 1.000000
## category_code11.57 0.000 1.000000
## category_code115.74 0.000 1.000000
## category_code118.72 0.000 1.000000
## category_code12.71 0.000 1.000000
## category_code12.73 0.000 1.000000
## category_code12.94 0.000 1.000000
## category_code120.98 0.000 1.000000
## category_code122.66 0.000 1.000000
## category_code13.66 0.000 1.000000
## category_code13.87 0.000 1.000000
## category_code13.89 0.000 1.000000
## category_code13.91 0.000 1.000000
## category_code132.41 0.000 1.000000
## category_code134.26 0.000 1.000000
## category_code138.87 0.000 1.000000
## category_code14.12 0.000 1.000000
## category_code14.14 0.000 1.000000
## category_code14.33 0.000 1.000000
## category_code148.13 0.000 1.000000
## category_code15.05 0.000 1.000000
## category_code15.28 0.000 1.000000
## category_code16.18 0.000 1.000000
## category_code16.20 0.000 1.000000
## category_code16.67 0.000 1.000000
## category_code162.04 0.000 1.000000
## category_code1666.64 0.000 1.000000
## category_code17.34 0.000 1.000000
## category_code17.36 0.000 1.000000
## category_code173.42 0.000 1.000000
## category_code18.03 0.000 1.000000
## category_code18.29 0.000 1.000000
## category_code18.50 0.000 1.000000
## category_code18.52 0.000 1.000000
## category_code18.98 0.000 1.000000
## category_code19.49 0.000 1.000000
## category_code19.68 0.000 1.000000
## category_code19.95 0.000 1.000000
## category_code2.29 0.000 1.000000
## category_code2.31 0.000 1.000000
## category_code2.52 0.000 1.000000
## category_code2.78 0.000 1.000000
## category_code2.99 0.000 1.000000
## category_code20.81 0.000 1.000000
## category_code20.83 0.000 1.000000
## category_code212.94 0.000 1.000000
## category_code219.88 0.000 1.000000
## category_code22.43 0.000 1.000000
## category_code22.66 0.000 1.000000
## category_code222.22 0.000 1.000000
## category_code23.13 0.000 1.000000
## category_code23.14 0.000 1.000000
## category_code23.15 0.000 1.000000
## category_code23.33 0.000 1.000000
## category_code234.94 0.000 1.000000
## category_code24.28 0.000 1.000000
## category_code245.93 0.000 1.000000
## category_code25.44 0.000 1.000000
## category_code25.46 0.000 1.000000
## category_code26.62 0.000 1.000000
## category_code27.08 0.000 1.000000
## category_code27.29 0.000 1.000000
## category_code27.34 0.000 1.000000
## category_code27.75 0.000 1.000000
## category_code27.78 0.000 1.000000
## category_code270.67 0.000 1.000000
## category_code28.94 0.000 1.000000
## category_code29.63 0.000 1.000000
## category_code3.45 0.000 1.000000
## category_code3.47 0.000 1.000000
## category_code3.89 0.000 1.000000
## category_code30.09 0.000 1.000000
## category_code300.90 0.000 1.000000
## category_code31.23 0.000 1.000000
## category_code314.79 0.000 1.000000
## category_code32.38 0.000 1.000000
## category_code33.33 0.000 1.000000
## category_code33.54 0.000 1.000000
## category_code34.70 0.000 1.000000
## category_code34.72 0.000 1.000000
## category_code35.39 0.000 1.000000
## category_code37.04 0.000 1.000000
## category_code370.35 0.000 1.000000
## category_code38.17 0.000 1.000000
## category_code39.33 0.000 1.000000
## category_code39.56 0.000 1.000000
## category_code4.14 0.000 1.000000
## category_code4.38 0.000 1.000000
## category_code4.61 0.000 1.000000
## category_code4.63 0.000 1.000000
## category_code40.49 0.000 1.000000
## category_code41.64 0.000 1.000000
## category_code41.67 0.000 1.000000
## category_code43.96 0.000 1.000000
## category_code46.27 0.000 1.000000
## category_code46.30 0.000 1.000000
## category_code472.02 0.000 1.000000
## category_code486.09 0.000 1.000000
## category_code5.30 0.000 1.000000
## category_code5.53 0.000 1.000000
## category_code5.56 0.000 1.000000
## category_code5.79 0.000 1.000000
## category_code50.90 0.000 1.000000
## category_code50.93 0.000 1.000000
## category_code53.22 0.000 1.000000
## category_code53.24 0.000 1.000000
## category_code532.38 0.000 1.000000
## category_code55.53 0.000 1.000000
## category_code555.30 0.000 1.000000
## category_code555.53 0.000 1.000000
## category_code57.87 0.000 1.000000
## category_code578.68 0.000 1.000000
## category_code6.02 0.000 1.000000
## category_code6.71 0.000 1.000000
## category_code6.92 0.000 1.000000
## category_code6.94 0.000 1.000000
## category_code64.79 0.000 1.000000
## category_code671.27 0.000 1.000000
## category_code69.42 0.000 1.000000
## category_code69.44 0.000 1.000000
## category_code7.85 0.000 1.000000
## category_code729.14 0.000 1.000000
## category_code74.07 0.000 1.000000
## category_code773.13 0.000 1.000000
## category_code78.68 0.000 1.000000
## category_code8.08 0.000 1.000000
## category_code8.10 0.000 1.000000
## category_code8.33 0.000 1.000000
## category_code8.54 0.000 1.000000
## category_code8.59 0.000 1.000000
## category_code8.77 0.000 1.000000
## category_code80.76 0.000 1.000000
## category_code83.31 0.000 1.000000
## category_code83.33 0.000 1.000000
## category_code87.94 0.000 1.000000
## category_code9.14 0.000 1.000000
## category_code9.24 0.000 1.000000
## category_code9.26 0.000 1.000000
## category_code92.57 0.000 1.000000
## category_code92.59 0.000 1.000000
## category_code97.20 0.000 1.000000
## category_codeaccessories.bag 1.386 0.165757
## category_codeaccessories.umbrella 0.766 0.443821
## category_codeapparel.glove 17.496 < 2e-16 ***
## category_codeapparel.shirt 2.111 0.034757 *
## category_codeapparel.sock NA NA
## category_codeapparel.trousers 1.445 0.148518
## category_codeapparel.tshirt 1.919 0.054928 .
## category_codeappliances.environment.air_conditioner 19.798 < 2e-16 ***
## category_codeappliances.environment.air_heater 4.037 5.43e-05 ***
## category_codeappliances.environment.climate -1.773 0.076196 .
## category_codeappliances.environment.fan 3.346 0.000819 ***
## category_codeappliances.environment.vacuum 5.064 4.12e-07 ***
## category_codeappliances.environment.water_heater 11.085 < 2e-16 ***
## category_codeappliances.iron 4.898 9.69e-07 ***
## category_codeappliances.ironing_board 4.034 5.49e-05 ***
## category_codeappliances.kitchen.blender 2.732 0.006295 **
## category_codeappliances.kitchen.coffee_grinder -2.240 0.025081 *
## category_codeappliances.kitchen.coffee_machine 1.186 0.235631
## category_codeappliances.kitchen.dishwasher 25.142 < 2e-16 ***
## category_codeappliances.kitchen.fryer 1.905 0.056847 .
## category_codeappliances.kitchen.grill 9.859 < 2e-16 ***
## category_codeappliances.kitchen.hood 8.781 < 2e-16 ***
## category_codeappliances.kitchen.juicer 5.460 4.78e-08 ***
## category_codeappliances.kitchen.kettle 1.961 0.049900 *
## category_codeappliances.kitchen.meat_grinder 4.810 1.51e-06 ***
## category_codeappliances.kitchen.microwave 5.556 2.76e-08 ***
## category_codeappliances.kitchen.mixer 0.755 0.450298
## category_codeappliances.kitchen.oven 20.983 < 2e-16 ***
## category_codeappliances.kitchen.refrigerators 35.402 < 2e-16 ***
## category_codeappliances.kitchen.steam_cooker 2.672 0.007538 **
## category_codeappliances.kitchen.toster -0.864 0.387710
## category_codeappliances.kitchen.washer 20.527 < 2e-16 ***
## category_codeappliances.personal.hair_cutter 1.432 0.152005
## category_codeappliances.personal.massager 2.442 0.014601 *
## category_codeappliances.personal.scales 2.131 0.033080 *
## category_codeappliances.sewing_machine 3.352 0.000801 ***
## category_codeappliances.steam_cleaner 0.600 0.548227
## category_codeauto.accessories.alarm NA NA
## category_codeauto.accessories.anti_freeze 0.296 0.767581
## category_codeauto.accessories.compressor 3.696 0.000219 ***
## category_codeauto.accessories.player NA NA
## category_codeauto.accessories.radar -0.794 0.427280
## category_codeauto.accessories.videoregister -0.076 0.939657
## category_codecomputers.components.cdrw 0.911 0.362302
## category_codecomputers.components.cooler -11.258 < 2e-16 ***
## category_codecomputers.components.cpu 10.652 < 2e-16 ***
## category_codecomputers.components.hdd 1.921 0.054765 .
## category_codecomputers.components.memory 2.122 0.033869 *
## category_codecomputers.components.motherboard -0.442 0.658831
## category_codecomputers.components.power_supply 3.765 0.000167 ***
## category_codecomputers.components.sound_card 0.597 0.550705
## category_codecomputers.components.videocards 8.180 2.87e-16 ***
## category_codecomputers.desktop 12.546 < 2e-16 ***
## category_codecomputers.ebooks 2.500 0.012434 *
## category_codecomputers.gaming 0.808 0.418840
## category_codecomputers.network.router -1.880 0.060136 .
## category_codecomputers.notebook 30.453 < 2e-16 ***
## category_codecomputers.peripherals.camera 2.419 0.015573 *
## category_codecomputers.peripherals.joystick -1.140 0.254338
## category_codecomputers.peripherals.keyboard 1.171 0.241568
## category_codecomputers.peripherals.monitor 6.350 2.17e-10 ***
## category_codecomputers.peripherals.mouse -0.193 0.847027
## category_codecomputers.peripherals.printer -5.374 7.72e-08 ***
## category_codecomputers.peripherals.scanner -3.224 0.001266 **
## category_codeconstruction.components.faucet 0.156 0.875816
## category_codeconstruction.tools.drill -0.144 0.885275
## category_codeconstruction.tools.generator 1.895 0.058077 .
## category_codeconstruction.tools.heater 0.491 0.623515
## category_codeconstruction.tools.light 0.087 0.930777
## category_codeconstruction.tools.pump 1.658 0.097230 .
## category_codeconstruction.tools.saw 0.710 0.477642
## category_codeconstruction.tools.screw -4.027 5.64e-05 ***
## category_codeconstruction.tools.welding 0.894 0.371199
## category_codecountry_yard.cultivator 2.201 0.027737 *
## category_codecountry_yard.lawn_mower 0.992 0.321283
## category_codecountry_yard.watering -0.776 0.437677
## category_codecountry_yard.weather_station 0.414 0.678707
## category_codeelectronics.audio.acoustic 11.260 < 2e-16 ***
## category_codeelectronics.audio.dictaphone -0.347 0.728616
## category_codeelectronics.audio.headphone -4.624 3.77e-06 ***
## category_codeelectronics.audio.microphone 5.647 1.64e-08 ***
## category_codeelectronics.audio.subwoofer 2.094 0.036237 *
## category_codeelectronics.camera.photo 8.874 < 2e-16 ***
## category_codeelectronics.camera.video 4.535 5.77e-06 ***
## category_codeelectronics.clocks 3.859 0.000114 ***
## category_codeelectronics.smartphone 15.087 < 2e-16 ***
## category_codeelectronics.tablet 10.224 < 2e-16 ***
## category_codeelectronics.telephone -0.383 0.701573
## category_codeelectronics.video.projector 1.261 0.207300
## category_codeelectronics.video.tv 32.850 < 2e-16 ***
## category_codefurniture.bathroom.bath 2.980 0.002885 **
## category_codefurniture.bathroom.toilet -0.982 0.326000
## category_codefurniture.bedroom.bed 1.849 0.064516 .
## category_codefurniture.bedroom.blanket 9.339 < 2e-16 ***
## category_codefurniture.bedroom.pillow 0.082 0.935020
## category_codefurniture.kitchen.chair 2.840 0.004514 **
## category_codefurniture.kitchen.table 2.253 0.024243 *
## category_codefurniture.living_room.cabinet 0.642 0.521019
## category_codefurniture.living_room.chair 4.811 1.51e-06 ***
## category_codefurniture.living_room.shelving 1.965 0.049366 *
## category_codefurniture.living_room.sofa 2.697 0.006989 **
## category_codekids.bottles 0.199 0.842217
## category_codekids.carriage 0.107 0.914859
## category_codekids.dolls 1.875 0.060812 .
## category_codekids.skates 17.142 < 2e-16 ***
## category_codekids.swing 0.946 0.344371
## category_codekids.toys 1.049 0.294225
## category_codemedicine.tools.tonometer 1.532 0.125641
## category_codesport.bicycle 1.310 0.190134
## category_codesport.tennis -1.019 0.308228
## category_codesport.trainer 3.299 0.000971 ***
## category_codestationery.battery 6.601 4.11e-11 ***
## category_codestationery.cartrige -15.107 < 2e-16 ***
## category_codestationery.paper -2.810 0.004956 **
## category_codestationery.stapler 0.091 0.927694
## category_codeUnknown 3.515 0.000439 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 142.2 on 117421 degrees of freedom
## Multiple R-squared: 0.6043, Adjusted R-squared: 0.6015
## F-statistic: 211 on 850 and 117421 DF, p-value: < 2.2e-16Interpretation
A Multiple Linear Regression model was developed using
brand and category_code as predictor variables
to estimate product price. By incorporating multiple predictors, the
model aims to explain more variation in product price and improve
prediction accuracy compared with the Simple Linear Regression
model.
5.6 Model Evaluation
#----------#
# Simple Model
#----------#
pred_simple <- predict(
simple_model,
newdata = test_data
)
valid_simple <- !is.na(pred_simple)
RMSE_simple <- sqrt(
mean((test_data$price[valid_simple] -
pred_simple[valid_simple])^2)
)
MAE_simple <- mean(
abs(test_data$price[valid_simple] -
pred_simple[valid_simple])
)
R2_simple <- 1 -
sum((test_data$price[valid_simple] -
pred_simple[valid_simple])^2) /
sum((test_data$price[valid_simple] -
mean(test_data$price[valid_simple]))^2)
#----------#
# Multiple Model
#----------#
pred_multiple <- predict(
multiple_model,
newdata = test_data
)
valid_multiple <- !is.na(pred_multiple)
RMSE_multiple <- sqrt(
mean((test_data$price[valid_multiple] -
pred_multiple[valid_multiple])^2)
)
MAE_multiple <- mean(
abs(test_data$price[valid_multiple] -
pred_multiple[valid_multiple])
)
R2_multiple <- 1 -
sum((test_data$price[valid_multiple] -
pred_multiple[valid_multiple])^2) /
sum((test_data$price[valid_multiple] -
mean(test_data$price[valid_multiple]))^2)
#----------#
# Comparison Table
#----------#
results <- data.frame(
Model = c("Simple Linear Regression",
"Multiple Linear Regression"),
RMSE = c(RMSE_simple,
RMSE_multiple),
MAE = c(MAE_simple,
MAE_multiple),
R2 = c(R2_simple,
R2_multiple)
)
resultsInterpretation
The model evaluation results show that the Multiple Linear Regression model outperformed the Simple Linear Regression model. The Multiple Linear Regression model achieved a lower RMSE (141.05 vs. 162.16) and lower MAE (63.15 vs. 77.31), indicating smaller prediction errors. It also obtained a higher R² value (0.6067 vs. 0.4801), meaning it explains approximately 60.67% of the variation in product price compared to 48.01% for the Simple Linear Regression model. Therefore, the Multiple Linear Regression model was selected as the final model because it provides better predictive performance and explanatory power.
5.7 Regression Diagnostic Plots
Diagnostic plots were generated to evaluate regression assumptions and model performance.
Interpretation
The diagnostic plots reveal patterns typical of right-skewed price data: residuals fan out at higher fitted values (heteroscedasticity) and the Q-Q plot departs from the line in the upper tail, driven by premium-priced outliers. The model remains useful for typical price ranges, but a log transformation of price would be a natural refinement if more precise prediction of expensive items were required.
5.8 Regression Modelling Summary
This study applied regression modelling techniques to predict product prices in e-commerce transactions. The findings showed that brand and product category significantly influence product price. Multiple linear regression produced better predictive performance than simple regression, indicating that combining multiple predictors improves price estimation accuracy.
6.0 Interpretation and Conclusion
This project set out to explore, clean, and model a large-scale e-commerce purchase history dataset, and all four objectives were met.
Data quality. The raw dataset contained duplicates
and substantial missing values in category_code,
brand, and user_id. These were resolved
through deduplication and “Unknown” placeholders, retaining the full
transaction history rather than discarding incomplete records.
Customer behaviour. EDA showed that transaction volume is concentrated in a small number of brands and categories, prices are heavily right-skewed, and purchasing activity follows clear hourly, weekly, and monthly rhythms. These patterns informed every modelling decision that followed.
Classification. Logistic Regression was the strongest classifier of high-value vs low-value purchases (ROC ≈ 0.91), with Random Forest a close second and Naive Bayes unusable due to its collapse into the majority class. Product category and brand were the dominant predictors, with temporal features adding modest value.
Regression. The multiple linear regression using brand and category explained roughly 60% of price variation, a substantial improvement over the brand-only baseline. The diagnostic plots suggest a log-price model as the most promising next step.
Practical implication. For an electronics retailer, the results point to a clear strategy: high-value purchases are driven primarily by what is bought (category) and from whom (premium brands such as Apple and Samsung), more than when. Marketing spend aimed at high-value customers should therefore target category and brand affinity first, with timing as a secondary lever.