WinningKaggleCompetition

#Libraries
library(xgboost) 
library(caret)
library(tictoc)

#Data
data <- read.csv(file.choose(), header = T)
data$Buy <- as.factor(data$Buy)
data <- data[,-1]

# Data Partition
set.seed(123)
ind <- sample(2, nrow(data), replace = T, prob = c(0.80, 0.20))
train <- data[ind==1,]
test <- data[ind==2,]

# Random search
modelLookup("xgbTree")
set.seed(123) 
cv <- trainControl(method="repeatedcv", 
                   number = 3, 
                   repeats = 2,
                   allowParallel = T,
                   verboseIter = T,
                   returnData = F,
                   search = 'random') 
set.seed(123) 
boo <- train(Buy ~ . , 
             data = train,  
             trControl = cv,
             method = "xgbTree",
             tuneLength = 3) 

#Confusion matrix
p <-  predict(boo, newdata = train, 'raw')
(tab <- table(Predicted = p, Actual = train$Buy))
p1 <-  predict(boo, newdata = test, 'raw')
(tab1 <- table(Predicted = p1, Actual = test$Buy))

# Submission
new <- read.csv(file.choose(), header = T)
p2 <-  predict(boo, newdata = new, 'prob')
p2 <- p2[,-1]
submission <- new$Id
submission <- data.frame(submission)
submission$Predicted <- p2
colnames(submission) <- c('Id', 'Predicted')
write.csv(submission, 'submission.csv', row.names=F)