forked from shettyavinita/DataScienceMidTermProject
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy path4.R
More file actions
37 lines (29 loc) · 1.24 KB
/
4.R
File metadata and controls
37 lines (29 loc) · 1.24 KB
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
##Load Training dataset
blogData_train <- read.csv("C:/Users/vanwu/Desktop/INFO 7390 ADS/Midterm/BlogFeedback/blogData_train.csv", header=FALSE)
##Load testing dataset
temp1 <- list.files(path="C:/Users/vanwu/Desktop/INFO 7390 ADS/Midterm/BlogFeedback/test/", pattern="*.csv")
blogData_test <-lapply(temp1, read.delim)
blogData_test <- read.csv(file = "C:/Users/vanwu/Desktop/INFO 7390 ADS/Midterm/BlogFeedback/test/blogData_test-2012.02.01.00_00.csv", header = FALSE)
#Training LR
lmFit <- lm(V281 ~ ., data = blogData_train)
summary(lmFit)
#Testing LR
rmseLm <- rmse(predict(lmFit,blogData_test),blogData_test$V281)
rmseLm
#Training CART
library(rpart)
cartFit <- rpart(V281 ~ ., method="anova", data=blogData_train)
#Visualize the tree
plot(cartFit, uniform=TRUE, main="Regression Tree")
text(cartFit, use.n=TRUE, all=TRUE, cex=.8)
#Test CART
rmseCart <-rmse(predict(cartFit,blogData_test),blogData_test$V281)
rmseCart
#Training Random forest
install.packages("randomForest")
library(randomForest)
rffit <- randomForest(V281 ~ ., data = blogData_train, ntree = 10, mtry = 10 ,importance=TRUE, na.action = na.omit)
summary(rffit)
#Test Random forest
rmseRF <-rmse(predict(rffit,blogData_test),blogData_test$V281)
rmseRF