Using devices such as Jawbone Up, Nike FuelBand, and Fitbit it is now possible to collect a large amount of data about personal activity relatively inexpensively. These type of devices are part of the quantified self movement – a group of enthusiasts who take measurements about themselves regularly to improve their health, to find patterns in their behavior, or because they are tech geeks. One thing that people regularly do is quantify how much of a particular activity they do, but they rarely quantify how well they do it.
In this project, your goal will be to use data from accelerometers on the belt, forearm, arm, and dumbell of 6 participants. They were asked to perform barbell lifts correctly and incorrectly in 5 different ways. More information is available from the website here: http://web.archive.org/web/20161224072740/http:/groupware.les.inf.puc-rio.br/har (see the section on the Weight Lifting Exercise Dataset). This project uses the h2o library for machine learning modeling.
training = read.csv("data/pml-training.csv")
# remove columns that will not provide value from training and testing data sets
nearZeroVariance <- nearZeroVar(training)
training_clean <- training[, -nearZeroVariance]
# check for columns that are 95% or more NA
nearAllNA <- sapply(training_clean, function(x) mean(is.na(x))) > 0.95
training <- training_clean[, nearAllNA==FALSE]
# remove columns without predictive value
train2 <- training[-c(1:7)]
# Split H2o data into Train/Validation/Test Sets
data_frame_h2o <- as.h2o(train2)
split_h2o <- h2o.splitFrame(data_frame_h2o, c(0.7, 0.15), seed = 1234)
train_h2o <- h2o.assign(split_h2o[[1]], "train") # 70%
valid_h2o <- h2o.assign(split_h2o[[2]], "valid") # 15%
test_h2o <- h2o.assign(split_h2o[[3]], "test") # 15%
y <- "classe" # Feature to Predict
x <- setdiff(names(data_frame_h2o), y) # All other data columns
h2o_gbm <- h2o.gbm(
model_id = 'h2o_gbm',
x = x,
y = y,
training_frame = train_h2o,
validation_frame = valid_h2o,
ntrees = 500,
max_depth = 6,
learn_rate = 0.1
)
h2o_rfe <- h2o.randomForest(
model_id = 'h2o_rfe',
x = x,
y = y,
training_frame = train_h2o,
validation_frame = valid_h2o,
ntrees = 500,
max_depth = 6
)
h2o_nb <- h2o.naiveBayes(
model_id = 'h2o_nb',
x = x,
y = y,
training_frame = train_h2o,
validation_frame = valid_h2o,
nfolds = 10,
laplace = 0
)
h2o_xgb <- h2o.xgboost(
x = x,
y = y,
training_frame = train_h2o,
validation_frame = valid_h2o
)
rsquare <-
data.frame(
Model = c(
"GBM",
"RFE",
"NB",
"XGBoost"
),
Accuracy = c(
h2o.r2(h2o_gbm),
h2o.r2(h2o_rfe),
h2o.r2(h2o_nb),
h2o.r2(h2o_xgb)
)
)
ggplot(rsquare, aes(x = reorder(Model, -Accuracy), y = Accuracy)) + geom_bar(stat = 'identity') +
theme_economist() + ggtitle('Comparison of Model R^2') + labs (x="Model sorted by R^2", y="Accuracy")
Print confusionmatrix
print(h2o.confusionMatrix(h2o_gbm, valid = TRUE))
## Confusion Matrix: Row labels: Actual class; Column labels: Predicted class
## A B C D E Error Rate
## A 816 1 0 0 0 0.0012 = 1 / 817
## B 1 574 0 0 0 0.0017 = 1 / 575
## C 0 2 527 1 0 0.0057 = 3 / 530
## D 0 0 6 469 0 0.0126 = 6 / 475
## E 0 0 0 2 541 0.0037 = 2 / 543
## Totals 817 577 533 472 541 0.0044 = 13 / 2,940
testing = read.csv("data/pml-testing.csv")
predictiondata <- as.h2o(testing)
predicted <- h2o.predict(object = h2o_gbm, newdata = predictiondata)
print(as.data.frame(predicted$predict))
## predict
## 1 B
## 2 A
## 3 B
## 4 A
## 5 A
## 6 E
## 7 D
## 8 B
## 9 A
## 10 A
## 11 B
## 12 C
## 13 B
## 14 A
## 15 E
## 16 E
## 17 A
## 18 B
## 19 B
## 20 B