1. Install dan memanggil package
packages <- c(
"tidyverse",
"readr",
"caret",
"randomForest",
"gbm",
"e1071",
"scales"
)
for (p in packages) {
if (!require(p, character.only = TRUE)) {
install.packages(p)
library(p, character.only = TRUE)
}
}
## Loading required package: tidyverse
## Warning: package 'tidyverse' was built under R version 4.5.3
## Warning: package 'ggplot2' was built under R version 4.5.3
## Warning: package 'tidyr' was built under R version 4.5.3
## Warning: package 'purrr' was built under R version 4.5.3
## Warning: package 'dplyr' was built under R version 4.5.3
## Warning: package 'forcats' was built under R version 4.5.1
## Warning: package 'lubridate' was built under R version 4.5.3
## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.2.1 ✔ readr 2.1.5
## ✔ forcats 1.0.1 ✔ stringr 1.5.1
## ✔ ggplot2 4.0.3 ✔ tibble 3.2.1
## ✔ lubridate 1.9.5 ✔ tidyr 1.3.2
## ✔ purrr 1.2.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
## Loading required package: caret
## Warning: package 'caret' was built under R version 4.5.3
## Loading required package: lattice
##
## Attaching package: 'caret'
##
## The following object is masked from 'package:purrr':
##
## lift
##
## Loading required package: randomForest
## Warning: package 'randomForest' was built under R version 4.5.3
## randomForest 4.7-1.2
## Type rfNews() to see new features/changes/bug fixes.
##
## Attaching package: 'randomForest'
##
## The following object is masked from 'package:dplyr':
##
## combine
##
## The following object is masked from 'package:ggplot2':
##
## margin
##
## Loading required package: gbm
## Warning: package 'gbm' was built under R version 4.5.3
## Loaded gbm 2.3.1
## This version of gbm is no longer under development. Consider transitioning to gbm3, https://github.com/gbm-developers/gbm3
## Loading required package: e1071
## Warning: package 'e1071' was built under R version 4.5.1
##
## Attaching package: 'e1071'
##
## The following object is masked from 'package:ggplot2':
##
## element
##
## Loading required package: scales
##
## Attaching package: 'scales'
##
## The following object is masked from 'package:purrr':
##
## discard
##
## The following object is masked from 'package:readr':
##
## col_factor
library(tidyverse)
library(readr)
library(caret)
library(randomForest)
library(gbm)
library(e1071)
library(scales)
set.seed(123)
2. Import dataset
datatugasmpml <- read_delim(
"C:/Users/Hype G12/Downloads/datatugasmpml.csv",
delim = ";",
show_col_types = FALSE
)
View(datatugasmpml)
dim(datatugasmpml)
## [1] 2392 15
names(datatugasmpml)
## [1] "StudentID" "Age" "Gender"
## [4] "Ethnicity" "ParentalEducation" "StudyTimeWeekly"
## [7] "Absences" "Tutoring" "ParentalSupport"
## [10] "Extracurricular" "Sports" "Music"
## [13] "Volunteering" "GPA" "GradeClass"
str(datatugasmpml)
## spc_tbl_ [2,392 × 15] (S3: spec_tbl_df/tbl_df/tbl/data.frame)
## $ StudentID : num [1:2392] 1001 1002 1003 1004 1005 ...
## $ Age : num [1:2392] 17 18 15 17 17 18 15 15 17 16 ...
## $ Gender : num [1:2392] 1 0 0 1 1 0 0 1 0 1 ...
## $ Ethnicity : num [1:2392] 0 0 2 0 0 0 1 1 0 0 ...
## $ ParentalEducation: num [1:2392] 2 1 3 3 2 1 1 4 0 1 ...
## $ StudyTimeWeekly : chr [1:2392] "19.833.722.807.854.700" "1.540.875.605.584.670" "421.056.976.881.226" "10.028.829.473.958.200" ...
## $ Absences : num [1:2392] 7 0 26 14 17 0 10 22 1 0 ...
## $ Tutoring : num [1:2392] 1 0 0 0 1 0 0 1 0 0 ...
## $ ParentalSupport : num [1:2392] 2 1 2 3 3 1 3 1 2 3 ...
## $ Extracurricular : num [1:2392] 0 0 0 1 0 1 0 1 0 1 ...
## $ Sports : num [1:2392] 0 0 0 0 0 0 1 0 1 0 ...
## $ Music : num [1:2392] 1 0 0 0 0 0 0 0 0 0 ...
## $ Volunteering : num [1:2392] 0 0 0 0 0 0 0 0 1 0 ...
## $ GPA : chr [1:2392] "2.929.195.591.667.680" "3.042.914.833.436.370" "0.1126022544661815" "20.542.181.397.029.400" ...
## $ GradeClass : num [1:2392] 2 1 4 3 4 1 2 4 2 0 ...
## - attr(*, "spec")=
## .. cols(
## .. StudentID = col_double(),
## .. Age = col_double(),
## .. Gender = col_double(),
## .. Ethnicity = col_double(),
## .. ParentalEducation = col_double(),
## .. StudyTimeWeekly = col_character(),
## .. Absences = col_double(),
## .. Tutoring = col_double(),
## .. ParentalSupport = col_double(),
## .. Extracurricular = col_double(),
## .. Sports = col_double(),
## .. Music = col_double(),
## .. Volunteering = col_double(),
## .. GPA = col_character(),
## .. GradeClass = col_double()
## .. )
## - attr(*, "problems")=<externalptr>
if (ncol(datatugasmpml) != 15) {
stop("Jumlah kolom tidak sesuai. Periksa kembali delimiter dataset.")
}
3. Pengecekan dan persiapan data
datatugasmpml <- distinct(datatugasmpml)
variabel_kategori <- c(
"Gender",
"Ethnicity",
"ParentalEducation",
"Tutoring",
"ParentalSupport",
"Extracurricular",
"Sports",
"Music",
"Volunteering"
)
datatugasmpml[variabel_kategori] <- lapply(
datatugasmpml[variabel_kategori],
as.factor
)
datatugasmpml$GradeClass <- factor(
datatugasmpml$GradeClass,
levels = 0:4
)
colSums(is.na(datatugasmpml))
## StudentID Age Gender Ethnicity
## 0 0 0 0
## ParentalEducation StudyTimeWeekly Absences Tutoring
## 0 0 0 0
## ParentalSupport Extracurricular Sports Music
## 0 0 0 0
## Volunteering GPA GradeClass
## 0 0 0
str(datatugasmpml)
## tibble [2,392 × 15] (S3: tbl_df/tbl/data.frame)
## $ StudentID : num [1:2392] 1001 1002 1003 1004 1005 ...
## $ Age : num [1:2392] 17 18 15 17 17 18 15 15 17 16 ...
## $ Gender : Factor w/ 2 levels "0","1": 2 1 1 2 2 1 1 2 1 2 ...
## $ Ethnicity : Factor w/ 4 levels "0","1","2","3": 1 1 3 1 1 1 2 2 1 1 ...
## $ ParentalEducation: Factor w/ 5 levels "0","1","2","3",..: 3 2 4 4 3 2 2 5 1 2 ...
## $ StudyTimeWeekly : chr [1:2392] "19.833.722.807.854.700" "1.540.875.605.584.670" "421.056.976.881.226" "10.028.829.473.958.200" ...
## $ Absences : num [1:2392] 7 0 26 14 17 0 10 22 1 0 ...
## $ Tutoring : Factor w/ 2 levels "0","1": 2 1 1 1 2 1 1 2 1 1 ...
## $ ParentalSupport : Factor w/ 5 levels "0","1","2","3",..: 3 2 3 4 4 2 4 2 3 4 ...
## $ Extracurricular : Factor w/ 2 levels "0","1": 1 1 1 2 1 2 1 2 1 2 ...
## $ Sports : Factor w/ 2 levels "0","1": 1 1 1 1 1 1 2 1 2 1 ...
## $ Music : Factor w/ 2 levels "0","1": 2 1 1 1 1 1 1 1 1 1 ...
## $ Volunteering : Factor w/ 2 levels "0","1": 1 1 1 1 1 1 1 1 2 1 ...
## $ GPA : chr [1:2392] "2.929.195.591.667.680" "3.042.914.833.436.370" "0.1126022544661815" "20.542.181.397.029.400" ...
## $ GradeClass : Factor w/ 5 levels "0","1","2","3",..: 3 2 5 4 5 2 3 5 3 1 ...
dim(datatugasmpml)
## [1] 2392 15
4. Visualisasi distribusi setiap variabel
tema_grafik <- theme_minimal(base_size = 12) +
theme(
plot.title = element_text(face = "bold", hjust = 0.5),
axis.title = element_text(face = "bold"),
axis.text = element_text(color = "black"),
panel.grid.minor = element_blank(),
legend.position = "bottom"
)
variabel_numerik <- names(datatugasmpml)[
sapply(datatugasmpml, is.numeric)
]
variabel_numerik <- setdiff(
variabel_numerik,
"StudentID"
)
for (var in variabel_numerik) {
p <- ggplot(
datatugasmpml,
aes(x = .data[[var]])
) +
geom_histogram(
fill = "skyblue",
color = "white",
bins = 25
) +
labs(
title = paste("Distribusi", var),
x = var,
y = "Frekuensi"
) +
tema_grafik
print(p)
}
for (var in variabel_kategori) {
p <- ggplot(
datatugasmpml,
aes(x = .data[[var]])
) +
geom_bar(
fill = "steelblue",
color = "white"
) +
coord_flip() +
labs(
title = paste("Distribusi", var),
x = var,
y = "Frekuensi"
) +
tema_grafik
print(p)
}
ggplot(
datatugasmpml,
aes(x = GradeClass, fill = GradeClass)
) +
geom_bar(color = "white") +
scale_fill_brewer(palette = "Set2") +
labs(
title = "Distribusi GradeClass",
x = "GradeClass",
y = "Frekuensi"
) +
tema_grafik
5. Visualisasi hubungan variabel dengan GradeClass
for (var in variabel_numerik) {
p <- ggplot(
datatugasmpml,
aes(
x = GradeClass,
y = .data[[var]],
fill = GradeClass
)
) +
geom_boxplot(alpha = 0.8) +
scale_fill_brewer(palette = "Set2") +
labs(
title = paste(var, "berdasarkan GradeClass"),
x = "GradeClass",
y = var
) +
tema_grafik
print(p)
}
for (var in variabel_kategori) {
p <- ggplot(
datatugasmpml,
aes(
x = .data[[var]],
fill = GradeClass
)
) +
geom_bar(position = "fill", color = "white") +
scale_fill_brewer(palette = "Set2") +
scale_y_continuous(
labels = percent_format()
) +
labs(
title = paste("Hubungan", var, "dengan GradeClass"),
x = var,
y = "Proporsi",
fill = "GradeClass"
) +
coord_flip() +
tema_grafik
print(p)
}
6. Statistik deskriptif
summary(datatugasmpml)
## StudentID Age Gender Ethnicity ParentalEducation
## Min. :1001 Min. :15.00 0:1170 0:1207 0:243
## 1st Qu.:1599 1st Qu.:15.00 1:1222 1: 493 1:728
## Median :2196 Median :16.00 2: 470 2:934
## Mean :2196 Mean :16.47 3: 222 3:367
## 3rd Qu.:2794 3rd Qu.:17.00 4:120
## Max. :3392 Max. :18.00
## StudyTimeWeekly Absences Tutoring ParentalSupport Extracurricular
## Length:2392 Min. : 0.00 0:1671 0:212 0:1475
## Class :character 1st Qu.: 7.00 1: 721 1:489 1: 917
## Mode :character Median :15.00 2:740
## Mean :14.54 3:697
## 3rd Qu.:22.00 4:254
## Max. :29.00
## Sports Music Volunteering GPA GradeClass
## 0:1666 0:1921 0:2016 Length:2392 0: 107
## 1: 726 1: 471 1: 376 Class :character 1: 269
## Mode :character 2: 391
## 3: 414
## 4:1211
##
data_numerik <- datatugasmpml %>%
select(where(is.numeric)) %>%
select(-any_of("StudentID"))
summary(data_numerik)
## Age Absences
## Min. :15.00 Min. : 0.00
## 1st Qu.:15.00 1st Qu.: 7.00
## Median :16.00 Median :15.00
## Mean :16.47 Mean :14.54
## 3rd Qu.:17.00 3rd Qu.:22.00
## Max. :18.00 Max. :29.00
for (var in variabel_kategori) {
cat("\nFrekuensi variabel:", var, "\n")
print(table(datatugasmpml[[var]]))
}
##
## Frekuensi variabel: Gender
##
## 0 1
## 1170 1222
##
## Frekuensi variabel: Ethnicity
##
## 0 1 2 3
## 1207 493 470 222
##
## Frekuensi variabel: ParentalEducation
##
## 0 1 2 3 4
## 243 728 934 367 120
##
## Frekuensi variabel: Tutoring
##
## 0 1
## 1671 721
##
## Frekuensi variabel: ParentalSupport
##
## 0 1 2 3 4
## 212 489 740 697 254
##
## Frekuensi variabel: Extracurricular
##
## 0 1
## 1475 917
##
## Frekuensi variabel: Sports
##
## 0 1
## 1666 726
##
## Frekuensi variabel: Music
##
## 0 1
## 1921 471
##
## Frekuensi variabel: Volunteering
##
## 0 1
## 2016 376
cat("\nDistribusi GradeClass:\n")
##
## Distribusi GradeClass:
print(table(datatugasmpml$GradeClass))
##
## 0 1 2 3 4
## 107 269 391 414 1211
7. Persiapan data untuk pemodelan
variabel_prediktor <- c(
"Age",
"Gender",
"Ethnicity",
"ParentalEducation",
"Absences",
"Tutoring",
"ParentalSupport",
"Extracurricular",
"Sports",
"Music",
"Volunteering"
)
data_model <- datatugasmpml %>%
select(all_of(variabel_prediktor), GradeClass) %>%
na.omit()
data_model <- droplevels(data_model)
dim(data_model)
## [1] 2392 12
str(data_model)
## tibble [2,392 × 12] (S3: tbl_df/tbl/data.frame)
## $ Age : num [1:2392] 17 18 15 17 17 18 15 15 17 16 ...
## $ Gender : Factor w/ 2 levels "0","1": 2 1 1 2 2 1 1 2 1 2 ...
## $ Ethnicity : Factor w/ 4 levels "0","1","2","3": 1 1 3 1 1 1 2 2 1 1 ...
## $ ParentalEducation: Factor w/ 5 levels "0","1","2","3",..: 3 2 4 4 3 2 2 5 1 2 ...
## $ Absences : num [1:2392] 7 0 26 14 17 0 10 22 1 0 ...
## $ Tutoring : Factor w/ 2 levels "0","1": 2 1 1 1 2 1 1 2 1 1 ...
## $ ParentalSupport : Factor w/ 5 levels "0","1","2","3",..: 3 2 3 4 4 2 4 2 3 4 ...
## $ Extracurricular : Factor w/ 2 levels "0","1": 1 1 1 2 1 2 1 2 1 2 ...
## $ Sports : Factor w/ 2 levels "0","1": 1 1 1 1 1 1 2 1 2 1 ...
## $ Music : Factor w/ 2 levels "0","1": 2 1 1 1 1 1 1 1 1 1 ...
## $ Volunteering : Factor w/ 2 levels "0","1": 1 1 1 1 1 1 1 1 2 1 ...
## $ GradeClass : Factor w/ 5 levels "0","1","2","3",..: 3 2 5 4 5 2 3 5 3 1 ...
cat("\nDistribusi GradeClass untuk pemodelan:\n")
##
## Distribusi GradeClass untuk pemodelan:
print(table(data_model$GradeClass))
##
## 0 1 2 3 4
## 107 269 391 414 1211
8. Fungsi evaluasi model
evaluasi_model <- function(prediksi, aktual) {
cm <- confusionMatrix(
prediksi,
aktual
)
tab <- cm$table
precision <- diag(tab) / rowSums(tab)
recall <- diag(tab) / colSums(tab)
precision[is.na(precision)] <- 0
recall[is.na(recall)] <- 0
f1 <- ifelse(
(precision + recall) == 0,
0,
2 * precision * recall / (precision + recall)
)
hasil <- data.frame(
Accuracy = as.numeric(cm$overall["Accuracy"]),
Macro_F1 = mean(f1)
)
return(hasil)
}
9. Pemodelan Random Forest, Gradient Boosting, dan SVM
proporsi_training <- c(0.50, 0.75, 0.80, 0.90)
hasil_semua <- data.frame()
kontrol_cv <- trainControl(
method = "cv",
number = 3,
allowParallel = FALSE
)
for (proporsi in proporsi_training) {
cat("\n====================================\n")
cat("Proporsi Training:", proporsi * 100, "%\n")
cat("====================================\n")
set.seed(123)
indeks <- createDataPartition(
data_model$GradeClass,
p = proporsi,
list = FALSE
)
data_train <- data_model[indeks, ]
data_test <- data_model[-indeks, ]
# Random Forest
grid_rf <- expand.grid(
mtry = c(2, 3, 4)
)
model_rf <- train(
GradeClass ~ .,
data = data_train,
method = "rf",
trControl = kontrol_cv,
tuneGrid = grid_rf,
ntree = 150,
importance = TRUE
)
prediksi_rf <- predict(
model_rf,
newdata = data_test
)
evaluasi_rf <- evaluasi_model(
prediksi_rf,
data_test$GradeClass
)
hasil_semua <- rbind(
hasil_semua,
data.frame(
Proporsi_Training = proporsi,
Model = "Random Forest",
Accuracy = evaluasi_rf$Accuracy,
Macro_F1 = evaluasi_rf$Macro_F1
)
)
# Gradient Boosting
grid_gbm <- expand.grid(
n.trees = c(100, 150),
interaction.depth = c(1, 2),
shrinkage = 0.05,
n.minobsinnode = 10
)
model_gbm <- train(
GradeClass ~ .,
data = data_train,
method = "gbm",
trControl = kontrol_cv,
tuneGrid = grid_gbm,
verbose = FALSE
)
prediksi_gbm <- predict(
model_gbm,
newdata = data_test
)
evaluasi_gbm <- evaluasi_model(
prediksi_gbm,
data_test$GradeClass
)
hasil_semua <- rbind(
hasil_semua,
data.frame(
Proporsi_Training = proporsi,
Model = "Gradient Boosting",
Accuracy = evaluasi_gbm$Accuracy,
Macro_F1 = evaluasi_gbm$Macro_F1
)
)
# Support Vector Machine
metode_svm <- c(
"svmLinear",
"svmPoly",
"svmRadial"
)
for (metode in metode_svm) {
model_svm <- train(
GradeClass ~ .,
data = data_train,
method = metode,
trControl = kontrol_cv,
preProcess = c("center", "scale"),
tuneLength = 2
)
prediksi_svm <- predict(
model_svm,
newdata = data_test
)
evaluasi_svm <- evaluasi_model(
prediksi_svm,
data_test$GradeClass
)
hasil_semua <- rbind(
hasil_semua,
data.frame(
Proporsi_Training = proporsi,
Model = paste("SVM", metode),
Accuracy = evaluasi_svm$Accuracy,
Macro_F1 = evaluasi_svm$Macro_F1
)
)
}
}
##
## ====================================
## Proporsi Training: 50 %
## ====================================
##
## ====================================
## Proporsi Training: 75 %
## ====================================
##
## ====================================
## Proporsi Training: 80 %
## ====================================
##
## ====================================
## Proporsi Training: 90 %
## ====================================
10. Hasil perbandingan seluruh model
hasil_semua <- hasil_semua %>%
arrange(
desc(Accuracy),
desc(Macro_F1)
)
print(hasil_semua)
## Proporsi_Training Model Accuracy Macro_F1
## 1 0.90 SVM svmLinear 0.7383966 0.5321094
## 2 0.75 Gradient Boosting 0.7294118 0.5846636
## 3 0.90 Random Forest 0.7215190 0.5093856
## 4 0.90 Gradient Boosting 0.7172996 0.5288865
## 5 0.75 Random Forest 0.7126050 0.5448835
## 6 0.75 SVM svmLinear 0.7126050 0.4953555
## 7 0.50 SVM svmLinear 0.7018425 0.4960243
## 8 0.80 Gradient Boosting 0.6953782 0.5341483
## 9 0.50 Gradient Boosting 0.6901173 0.5503415
## 10 0.80 SVM svmLinear 0.6890756 0.4698011
## 11 0.80 Random Forest 0.6806723 0.4927170
## 12 0.50 Random Forest 0.6608040 0.4793023
## 13 0.90 SVM svmPoly 0.6540084 0.3486229
## 14 0.75 SVM svmPoly 0.6537815 0.3468657
## 15 0.75 SVM svmRadial 0.6504202 0.3600123
## 16 0.90 SVM svmRadial 0.6497890 0.3563081
## 17 0.80 SVM svmPoly 0.6386555 0.3314731
## 18 0.80 SVM svmRadial 0.6323529 0.3330138
## 19 0.50 SVM svmPoly 0.6314908 0.3218107
## 20 0.50 SVM svmRadial 0.6180905 0.3163835
11. Menentukan model terbaik
model_terbaik <- hasil_semua %>%
arrange(
desc(Accuracy),
desc(Macro_F1)
) %>%
slice(1)
cat("\nModel terbaik:\n")
##
## Model terbaik:
print(model_terbaik)
## Proporsi_Training Model Accuracy Macro_F1
## 1 0.9 SVM svmLinear 0.7383966 0.5321094
12. Visualisasi perbandingan Accuracy dan Macro F1
# Grafik Accuracy
ggplot(
hasil_semua,
aes(
x = factor(Proporsi_Training),
y = Accuracy,
fill = Model
)
) +
geom_col(
position = position_dodge(width = 0.85),
width = 0.7
) +
geom_text(
aes(label = percent(Accuracy, accuracy = 0.1)),
position = position_dodge(width = 0.85),
vjust = -0.5,
size = 2.8
) +
scale_y_continuous(
labels = percent_format(),
limits = c(0, 1.12),
breaks = seq(0, 1, 0.2)
) +
labs(
title = "Perbandingan Accuracy Model",
x = "Proporsi Data Training",
y = "Accuracy",
fill = "Model"
) +
tema_grafik
# Grafik Macro F1
ggplot(
hasil_semua,
aes(
x = factor(Proporsi_Training),
y = Macro_F1,
fill = Model
)
) +
geom_col(
position = position_dodge(width = 0.85),
width = 0.7
) +
geom_text(
aes(label = percent(Macro_F1, accuracy = 0.1)),
position = position_dodge(width = 0.85),
vjust = -0.5,
size = 2.8
) +
scale_y_continuous(
labels = percent_format(),
limits = c(0, 1.12),
breaks = seq(0, 1, 0.2)
) +
labs(
title = "Perbandingan Macro F1 Model",
x = "Proporsi Data Training",
y = "Macro F1",
fill = "Model"
) +
tema_grafik
13. Analisis Random Forest
set.seed(123)
indeks_80 <- createDataPartition(
data_model$GradeClass,
p = 0.80,
list = FALSE
)
data_train_80 <- data_model[indeks_80, ]
data_test_80 <- data_model[-indeks_80, ]
model_rf_80 <- randomForest(
GradeClass ~ .,
data = data_train_80,
ntree = 150,
importance = TRUE
)
prediksi_rf_80 <- predict(
model_rf_80,
newdata = data_test_80
)
confusionMatrix(
prediksi_rf_80,
data_test_80$GradeClass
)
## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1 2 3 4
## 0 4 2 0 0 0
## 1 10 25 10 1 1
## 2 6 19 36 15 0
## 3 0 4 23 42 18
## 4 1 3 9 24 223
##
## Overall Statistics
##
## Accuracy : 0.6933
## 95% CI : (0.6497, 0.7344)
## No Information Rate : 0.5084
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.5304
##
## Mcnemar's Test P-Value : NA
##
## Statistics by Class:
##
## Class: 0 Class: 1 Class: 2 Class: 3 Class: 4
## Sensitivity 0.190476 0.47170 0.46154 0.51220 0.9215
## Specificity 0.995604 0.94799 0.89950 0.88579 0.8419
## Pos Pred Value 0.666667 0.53191 0.47368 0.48276 0.8577
## Neg Pred Value 0.963830 0.93473 0.89500 0.89717 0.9120
## Prevalence 0.044118 0.11134 0.16387 0.17227 0.5084
## Detection Rate 0.008403 0.05252 0.07563 0.08824 0.4685
## Detection Prevalence 0.012605 0.09874 0.15966 0.18277 0.5462
## Balanced Accuracy 0.593040 0.70984 0.68052 0.69899 0.8817
varImpPlot(
model_rf_80,
main = "Variable Importance Random Forest"
)
14. Analisis Gradient Boosting
set.seed(123)
model_gbm_raw <- gbm(
GradeClass ~ .,
data = data_train_80,
distribution = "multinomial",
n.trees = 150,
interaction.depth = 2,
shrinkage = 0.05,
bag.fraction = 0.8,
n.minobsinnode = 10,
cv.folds = 3,
verbose = FALSE
)
jumlah_pohon <- gbm.perf(
model_gbm_raw,
method = "cv",
plot.it = TRUE
)
importance_gbm <- summary(
model_gbm_raw,
n.trees = jumlah_pohon,
plotit = TRUE
)
print(importance_gbm)
## var rel.inf
## Absences Absences 87.9258164
## ParentalSupport ParentalSupport 5.9277324
## Tutoring Tutoring 2.6765965
## Sports Sports 0.9406630
## Extracurricular Extracurricular 0.6823206
## Music Music 0.5157937
## Ethnicity Ethnicity 0.4455832
## ParentalEducation ParentalEducation 0.3365195
## Age Age 0.1965950
## Volunteering Volunteering 0.1793017
## Gender Gender 0.1730781
15. Perbandingan kernel SVM
hasil_svm <- hasil_semua %>%
filter(grepl("^SVM", Model)) %>%
group_by(Model) %>%
slice_max(
order_by = Accuracy,
n = 1,
with_ties = FALSE
) %>%
ungroup() %>%
mutate(
Kernel = sub("^SVM svm", "", Model),
Kernel = tools::toTitleCase(Kernel)
)
print(hasil_svm)
## # A tibble: 3 × 5
## Proporsi_Training Model Accuracy Macro_F1 Kernel
## <dbl> <chr> <dbl> <dbl> <chr>
## 1 0.9 SVM svmLinear 0.738 0.532 Linear
## 2 0.9 SVM svmPoly 0.654 0.349 Poly
## 3 0.75 SVM svmRadial 0.650 0.360 Radial
ggplot(
hasil_svm,
aes(
x = reorder(Kernel, Accuracy),
y = Accuracy,
fill = Kernel
)
) +
geom_col(
width = 0.65,
show.legend = FALSE
) +
geom_text(
aes(label = percent(Accuracy, accuracy = 0.1)),
hjust = -0.15,
size = 3.5
) +
coord_flip() +
scale_y_continuous(
labels = percent_format(),
limits = c(0, 1.12)
) +
labs(
title = "Perbandingan Accuracy Kernel SVM",
x = "Kernel SVM",
y = "Accuracy"
) +
tema_grafik
