crossvalidation.Rmd

---
title: "R Notebook"
output: html_notebook
---


```{r}
library(neuralnet)
library(boot)
set.seed(12345)

df <- read.csv("dataset.csv")
#apply(df,2,function(x) sum(is.na(x)))

#str(df)

mydata <- df[c(2,4:6,11:14)] 
#apply(mydata,2,function(x) sum(is.na(x)))

lm.fit <- glm(Default~.,data=mydata)
cv.glm(mydata,lm.fit,K=10)$delta[1]
```
```{r}
set.seed(450)

cv.error <- NULL
k <- 10
 
library(plyr)

pbar<-create_progress_bar('text')
pbar$init(k)

for(i in 1:k){
    index <- sample(1:nrow(mydata),round(0.9*nrow(mydata)))
    train.cv <- scaled[index,]
    test.cv <- scaled[-index,]
    
    nn <- neuralnet(f,data=train.cv,hidden=c(5,2),linear.output=F,stepmax = 10000000)
    
    pr.nn <- compute(nn,mydata[,1:7])
    pr.nn <- pr.nn$net.result*(max(mydata$Default)-min(mydata$Default))+min(mydata$Default)
    
    test.cv.r <- (df$Default)*(max(mydata$Default)-min(mydata$Default))+min(mydata$Default)
    
    cv.error[i] <- sum((test.cv.r - pr.nn)^2)/nrow(df)
    
    pbar$step()
}


```
```{r}
mean(cv.error)

boxplot(cv.error,xlab='MSE CV',col='cyan',
        border='blue',names='CV error (MSE)',
        main='CV error (MSE) for NN',horizontal=TRUE)
```