MachineLearning/DecisionTree.R at main · a-pt/MachineLearning · GitHub

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
#Decision Regression Tree

movie <- read.csv("D:/Online courses/ML(Rstudio)/resources/Complete ML in R/3. Decision Trees/Movie_regression.csv",header=TRUE)
View(movie)

#Prepocessing

summary(movie)
plot(movie$Collection,movie$Marketing.expense)
mean <- mean(movie$Time_taken,na.rm=TRUE)
movie$Time_taken[is.na(movie$Time_taken)] <- mean

#Test Train split

install.packages("caTools")
library(caTools)
set.seed(0)
split <- sample.split(movie,SplitRatio = 0.8)
train_set <- subset(movie,split==TRUE)
test_set <- subset(movie,split==FALSE)


#install packages

install.packages('rpart')
install.packages("rpart.plot")
library(rpart)
library(rpart.plot)

?rpart

regtree <- rpart(formula=Collection~.,data=train_set,control=rpart.control(maxdepth=3))
View(regtree)

rpart.plot(regtree,box.palette="RdBu",digits=-3)

#prediction
test_set$predict <- predict(regtree,test_set,type="vector")

#performance
mse2 <- mean((test_set$predict-test_set$Collection)^2)

#pruning
fulltree <- rpart(formula=Collection~.,data=train_set,control=rpart.control(cp=0))
rpart.plot(fulltree,box.palette="RdBu",digits=-3)

printcp(fulltree)
plotcp(regtree)
index<-which.min(regtree$cptable[,"xerror"])
min_cp <-regtree$cptable[index,"CP"]
prunedtree <- prune(fulltree,cp=min_cp)
rpart.plot(prunedtree,box.palette="RdBu",digits=-3)

#comparison(full,purned,reg)

test_set$full_predict <- predict(fulltree,test_set,type="vector")
full_mse2 <- mean((test_set$full_predict-test_set$Collection)^2)

test_set$pur_predict <- predict(prunedtree,test_set,type="vector")
prune_mse2 <- mean((test_set$pur_predict-test_set$Collection)^2)


#Classification Decision Tree

df <- read.csv("D:/Online courses/ML(Rstudio)/resources/Complete ML in R/3. Decision Trees/Movie_classification.csv",header=TRUE)
View(df)

summary(df)
df$Time_taken[is.na(df$Time_taken)] <- mean(df$Time_taken,na.rm=TRUE)
summary(df)

install.packages("caTools")
library(caTools)
set.seed(0)
split_c <-sample.split(df,SplitRatio = 0.8)
train_set_c <- subset(df,split_c==TRUE)
test_set_c <-subset(df,split_c==FALSE)

#install.packages("rpart")
#install.packages("rpart.plot")

classtree<- rpart(formula = Start_Tech_Oscar~.,data=train_set_c,method = "class",control=rpart.control(maxdepth=3))
rpart.plot(classtree,box.palette = "RdBu",digits=-3)

test_set_c$predict <- predict(classtree,test_set_c,type="class")
View(test_set_c)
table(test_set_c$predict,test_set_c$Start_Tech_Oscar)
64/107

#Bagging (Regression)
install.packages("randomForest")
library(randomForest)
#Do test train spilt,seed set
bagging <- randomForest(formula=Collection~.,data=train_set,mtry=17)
summary(bagging)

test_set$bagging_pred <- predict(bagging,test_set)
mse2_bagging<-mean((test_set$bagging_pred-test_set$Collection)^2)


#Random Forest (Regression)

#install.packages("randomForest")
#library(randomForest)

randomforest <- randomForest(Collection~.,train_set,ntree=500)
test_set$rfpredict <- predict(randomforest,test_set)
mse2_rf <- mean((test_set$rfpredict-test_set$Collection)^2)

#Gradient Boost
install.packages("gbm")
library(gbm)

#Preprocessing
#install.packages("dummies")
#library(dummies)
df2 <- movie
df2<-dummy.data.frame(df2)
df2 <- df2[,-12]
df2 <- df2[,-15]
set.seed(0)
split <- sample.split(df2,SplitRatio = 0.8)
train_set_gb <- subset(df2,split==TRUE)
test_set_gb <- subset(df2,split==FALSE)
gboosting <- gbm(Collection~.,data=train_set_gb,distribution = "gaussian", n.trees = 5000,interaction.depth = 4,shrinkage = 0.2,verbose = F)
summary(gboosting)
test_set_gb$predict <- predict(gboosting,test_set_gb,n.trees = 5000)
gboost_mse2 <- mean((test_set_gb$predict-test_set_gb$Collection)^2)


#AdaBoost

install.packages("adabag")
library(adabag)
train_set_c$Start_Tech_Oscar<-as.factor(train_set_c$Start_Tech_Oscar)
adaboost <- boosting(Start_Tech_Oscar~.,train_set_c,boos=TRUE)
ada_predict <- predict(adaboost,test_set_c)
table(ada_predict$class,test_set_c$Start_Tech_Oscar)

t1<-adaboost$trees[[1]]
plot(t1)
text(t1,pretty=100)

#XG Boost

install.packages("xgboost")
library(xgboost)

#Dmatrix format

train_y <- train_set_c$Start_Tech_Oscar=='1'
train_x <- model.matrix(Start_Tech_Oscar~.,-1,data=train_set_c)
train_x <- train_x[,-1]

test_y <- test_set_c$Start_Tech_Oscar=='1'
test_x <- model.matrix(Start_Tech_Oscar~.,-1,data=test_set_c)
test_x <- test_x[,-22]
test_x <- test_x[,-1]
Xmatrix_train <- xgb.DMatrix(data=train_x,label=train_y)
Xmatrix_test <- xgb.DMatrix(data=test_x,label=test_y)

xgboosting <- xgboost(data=Xmatrix_train,nrounds=50,objective="multi:softmax",eta=0.3,num_class=2,max_depth=10)
xgpred <- predict(xgboosting,Xmatrix_test)
table(xgpred,test_y)
68/107