R Statistics Cookbook

How to do it...

In this exercise, we will use gradient boosting with a real dataset containing wine types. Here we will have three classes, and they are slightly imbalanced:

1. We first load the modeling dataset and prepare the modeling data:

library(MASS)library(PRROC)library(precrec)set.seed(10)data = read.csv("./approved.csv")data = data[,-c(1,7)]data$Approved_ = "not_approved"data$Approved_[data$Approved == 1] <- "approved"data$Approved_ = as.factor(data$Approved_)data = data[,-1]

2. We split the data into training and testing groups:

trainIndex       <- createDataPartition(data$Approved_, p = .75,  list = FALSE,  times = 1) traindata        <- data[trainIndex,] testdata         <- data[-trainIndex,] 

3. We then fit our first model; let's call it baseline ...