-
Notifications
You must be signed in to change notification settings - Fork 0
/
ensembles.methods.r
246 lines (185 loc) · 7.07 KB
/
ensembles.methods.r
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
##### MADE BY ######
# 07/01/20
#Created by: Paraskevi Sifnaiou
#####################
############# Step 1: Data import ##############
setwd("C:/Users/sifne/Desktop/")
# Load libraries
library(readxl)
library(tidyverse)
library(caretEnsemble)
library(magrittr)
library(plyr)
library(caret)
library(ROCR)
online_shoppers_intention <- read.csv('online_shoppers_intention.csv')
boxplot(online_shoppers_intention)
onl_shop<-online_shoppers_intention #to preserve the original dataset
#Remove columns:11-17 to get higher accuracy
onl_shop[11:17] <- NULL
boxplot(onl_shop)
#check of NA's
sum(is.na(onl_shop))
summary(onl_shop)
str(onl_shop)
#Preprocess
onl_shop$Revenue <- make.names(onl_shop$Revenue)
onl_shop$Revenue<-as.factor(onl_shop$Revenue)
glimpse(onl_shop)
############# Data Split ###############
#Create index to split
index <- createDataPartition(onl_shop$Revenue,p=0.75,list=FALSE)
# Subset training set with index
onl_shop.training<-onl_shop[index,]
# Subset test set with index
onl_shop.test<-onl_shop[-index,]
##repeated k-fold cross validation
# define training control
control <- trainControl(method="repeatedcv", number=10, repeats=3,
savePredictions=TRUE, classProbs=TRUE,preProc = c("center","scale"))
####################### bagging algorithm #############################
seed <- 7
metric <- "Accuracy"
# Treebag
set.seed(seed)
fit.treebag <- train(Revenue~., data=onl_shop.training, method="treebag", metric=metric, trControl=control)
# Random Forest
set.seed(seed)
fit.rf <- train(Revenue~., data=onl_shop.training, method="rf", metric=metric, trControl=control)
#treebag testing set accuracy
predictions_treebag<-predict(object=fit.treebag ,onl_shop.test, type="raw")
table(predictions_treebag)
confusionMatrix(predictions_treebag,onl_shop.test$Revenue)
#random forest testing set accuracy
predictions_rf<-predict(object=fit.rf ,onl_shop.test, type="raw")
table_rf<-table(predictions_rf)
table_rf
confusionMatrix(predictions_rf,onl_shop.test$Revenue)
# summarize results
bagging_results <- resamples(list(treebag=fit.treebag, rf=fit.rf))
summary(bagging_results)
dotplot(bagging_results)
######################### stacking algorithm ###################################
# create submodels
algorithmList <- c( 'rpart', 'knn' ,'nb')
set.seed(seed)
models <- caretList(Revenue~., data=onl_shop.training, trControl=control, methodList=algorithmList)
#results
results <- resamples(models)
summary(results)
dotplot(results)
# correlation between results
modelCor(results)
splom(results)
################## Stack 2 models with best correlations: only knn & nb ############
#knn
set.seed(seed)
stack.knn <- caretStack(models, method="knn", metric="Accuracy", trControl=control)
print(stack.knn)
prediction_knn2<-predict(stack.knn ,onl_shop.test)
table(prediction_knn2)
confusionMatrix(prediction_knn2,onl_shop.test$Revenue)
#nb
set.seed(seed)
stack.nb <- caretStack(models, method="nb", metric="Accuracy", trControl=control)
print(stack.nb)
prediction_nb2<-predict(stack.nb ,onl_shop.test)
table(prediction_nb2)
confusionMatrix(prediction_nb2,onl_shop.test$Revenue)
###################### Measure Performance ##############################
####random forest#####
#AUC
pred_rf_raw <- predict(fit.rf,onl_shop.test,type = "raw")
prediction_raw <- prediction(as.numeric(predictions_rf), onl_shop.test$Revenue)
tpr_fpr <- performance(prediction_raw,"tpr","fpr")
plot(tpr_fpr)
plot(tpr_fpr,colorize=TRUE,main = "ROC Curve",
ylab = "sensivity",
xlab = "specifity")
abline(a= 0,b=1)
pred_auc <- performance(prediction_raw ,measure="auc")
pred_auc
auc_value <- [email protected][[1]]
auc_value <- round(auc_value,4)
legend(.6,.4,auc_value,title = "AUC")
######treebag########
#AUC
pred_rf_raw <- predict(fit.treebag,onl_shop.test,type = "raw")
prediction_raw <- prediction(as.numeric(predictions_treebag), onl_shop.test$Revenue)
tpr_fpr <- performance(prediction_raw,"tpr","fpr")
plot(tpr_fpr,colorize=TRUE,main = "ROC Curve",
ylab = "sensivity",
xlab = "specifity")
abline(a= 0,b=1)
pred_auc <- performance(prediction_raw ,measure="auc")
pred_auc
auc_value <- [email protected][[1]]
auc_value <- round(auc_value,4)
legend(.6,.4,auc_value,title = "AUC")
########knn############
set.seed(seed)
pred_knn_prob<- predict(stack.knn,onl_shop.test,type = "prob")
prediction_knn <- prediction( pred_knn_prob, onl_shop.test$Revenue)
#evalutation of accuracy
accuracy <- performance(prediction_knn,"acc")
plot(accuracy)
#precision vs recall
prec_vs_recall <- performance(prediction_knn,"prec","rec")
plot(prec_vs_recall)
#AUC
pred_knn_raw <- predict(stack.knn,onl_shop.test,type = "raw")
prediction_raw <- prediction(as.numeric(pred_knn_raw), onl_shop.test$Revenue)
tpr_fpr <- performance(prediction_raw,"tpr","fpr")
plot(tpr_fpr)
plot(tpr_fpr,colorize=TRUE,main = "ROC Curve",
ylab = "sensivity",
xlab = "specifity")
abline(a= 0,b=1)
pred_auc <- performance(prediction_raw ,measure="auc")
pred_auc
auc_value <- [email protected][[1]]
auc_value <- round(auc_value,4)
legend(.6,.4,auc_value,title = "AUC")
#########nb###########
set.seed(seed)
pred_nb_prob<- predict(stack.nb,onl_shop.test,type = "prob")
prediction_nb <- prediction( pred_nb_prob, onl_shop.test$Revenue)
#evalutation of accuracy
accuracy <- performance(prediction_nb,"acc")
plot(accuracy)
#precision vs recall
prec_vs_recall <- performance(prediction_nb,"prec","rec")
plot(prec_vs_recall)
#AUC
pred_nb_raw <- predict(stack.nb,onl_shop.test,type = "raw")
prediction_raw <- prediction(as.numeric(pred_nb_raw), onl_shop.test$Revenue)
tpr_fpr <- performance(prediction_raw,"tpr","fpr")
plot(tpr_fpr)
plot(tpr_fpr,colorize=TRUE,main = "ROC Curve",
ylab = "sensivity",
xlab = "specifity")
abline(a= 0,b=1)
pred_auc <- performance(prediction_raw ,measure="auc")
pred_auc
auc_value <- [email protected][[1]]
auc_value <- round(auc_value,4)
legend(.6,.4,auc_value,title = "AUC")
#############training and testing time#################
#bagging
##treebag: training time
system.time(fit.treebag <- train(Revenue~., data=onl_shop.training, method="treebag", metric=metric, trControl=control))
##treebag: testing time
system.time(predictions_treebag<-predict(object=fit.treebag ,onl_shop.test, type="raw"))
##random forest: training time
system.time(fit.rf <- train(Revenue~., data=onl_shop.training, method="rf", metric=metric, trControl=control))
##random forest: testing time
system.time(predictions_rf<-predict(object=fit.rf ,onl_shop.test, type="raw"))
#stacking
#training knn
system.time(stack.knn <- caretStack(models, method="knn", metric="Accuracy", trControl=control))
#testing knn
system.time(prediction_knn2<-predict(stack.knn ,onl_shop.test))
#training nb
system.time(stack.nb <- caretStack(models, method="nb", metric="Accuracy", trControl=control))
#testing nb
system.time(pred_nb_prob<- predict(stack.nb,onl_shop.test,type = "prob"))