#Dealing with extremely low proporation of one of classes
setwd("~/Documents/NYDSA/DATA SET/DemoWork/Thyroidism-RareEvent")
library(caret); library(pROC); library(DMwR);
## Warning: package 'caret' was built under R version 3.1.2
## Loading required package: lattice
## Warning: package 'lattice' was built under R version 3.1.2
## Loading required package: ggplot2
## Type 'citation("pROC")' for a citation.
##
## Attaching package: 'pROC'
##
## The following objects are masked from 'package:stats':
##
## cov, smooth, var
##
## Loading required package: grid
hyper <-read.csv('http://archive.ics.uci.edu/ml/machine-learning-databases/thyroid-disease/hypothyroid.data', header=F)
names <- read.csv('http://archive.ics.uci.edu/ml/machine-learning-databases/thyroid-disease/hypothyroid.names', header=F, sep='\t')[[1]]
names <- gsub(pattern =":|[.]",x = names, replacement="")
(colnames(hyper) <- names)
## [1] "hypothyroid, negative" "age"
## [3] "sex" "on_thyroxine"
## [5] "query_on_thyroxine" "on_antithyroid_medication"
## [7] "thyroid_surgery" "query_hypothyroid"
## [9] "query_hyperthyroid" "pregnant"
## [11] "sick" "tumor"
## [13] "lithium" "goitre"
## [15] "TSH_measured" "TSH"
## [17] "T3_measured" "T3"
## [19] "TT4_measured" "TT4"
## [21] "T4U_measured" "T4U"
## [23] "FTI_measured" "FTI"
## [25] "TBG_measured" "TBG"
#data cleaning:
#1. Change column name from "hypothyroid, negative" to "target"
colnames(hyper) <-c("target", "age", "sex", "on_thyroxine", "query_on_thyroxine", "on_antithyroid_medication", "thyroid_surgery", "query_hypothyroid", "query_hyperthyroid", "pregnant", "sick", "tumor", "lithium","goitre", "TSH_measured", "TSH", "T3_measured", "T3", "TT4_measured","TT4", "T4U_measured", "T4U", "FTI_measured", "FTI", "TBG_measured","TBG")
hyper$target <- ifelse(hyper$target=='negative',0,1)
(prop.table(table(hyper$target))) #skewness
##
## 0 1
## 0.95226051 0.04773949
#2. binarize characters
tmp <- sapply(hyper, is.factor)
hyper [tmp] <- lapply(hyper[tmp], as.character)
hyper[ hyper == "?" ] = NA
hyper[ hyper == "f" ] = 0; hyper[ hyper == "t" ] = 1
hyper[ hyper == "n" ] = 0; hyper[ hyper == "y" ] = 1
hyper[ hyper == "M" ] = 0; hyper[ hyper == "F" ] = 1
hyper[tmp] <- lapply(hyper[tmp], as.numeric)
#3. replace NA with mean of the column
repalceNAsWithMean <- function(x) {replace(x, is.na(x), mean(x[!is.na(x)]))}
hyper <- repalceNAsWithMean(hyper)
#model preparation
set.seed(100)
inTrain <- createDataPartition(hyper$target, p=.5, list=FALSE)
train <- hyper[inTrain,]; test <- hyper[-inTrain,]
#model: treebag
ctrl <- trainControl(method="repeatedcv", repeats=3)
bagModel <- train(target ~., data=train, method="treebag", trControl = ctrl)
## Loading required package: ipred
## Warning: package 'ipred' was built under R version 3.1.2
## Loading required package: plyr
##
## Attaching package: 'plyr'
##
## The following object is masked from 'package:DMwR':
##
## join
predictors <- names(train[2:ncol(train)]) #Xi
Predicts <- predict(bagModel$finalModel, test[, predictors])
#model evaluation: pROC
auc <- roc(test$target, Predicts)
auc[9]
## $auc
## Area under the curve: 0.9915
plot(auc,ylim=c(0,1), print.thres=TRUE, main=paste('AUC:', auc[9]))
##
## Call:
## roc.default(response = test$target, predictor = Predicts)
##
## Data: Predicts in 1504 controls (test$target 0) < 77 cases (test$target 1).
## Area under the curve: 0.9915
abline(h=1, col='blue', lwd=3)
abline(h=0, col='red', lwd=3)
#Now SMOTE:
#The SMOTE function oversamples your rare event by using bootstrapping and k-nearest neighbor to synthetically create additional observations of that event. The definition of rare event is usually attributed to any outcome/dependent/target/response variable that happens less than 15% of the time
train$target <- as.factor(train$target)
train.SMOTE <- SMOTE(target ~ ., train, perc.over = 100, perc.under=200) #perc over (increase percent) & under (decrease percentage)
#post SMOTE vs. before SMOTE
train.SMOTE[,1]; train[,1]
## [1] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [36] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [71] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [106] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [141] 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## [176] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## [211] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## [246] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## [281] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## Levels: 0 1
## [1] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## [35] 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
## [69] 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [103] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [137] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [171] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [205] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [239] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [273] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [307] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [341] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [375] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [409] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [443] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [477] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [511] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [545] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [579] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [613] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [647] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [681] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [715] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [749] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [783] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [817] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [851] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [885] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [919] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [953] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [987] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1021] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1055] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1089] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1123] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1157] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1191] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1225] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1259] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1293] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1327] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1361] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1395] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1429] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1463] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1497] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1531] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## [1565] 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
## Levels: 0 1
(postSMOTE <- prop.table(table(train.SMOTE$target))); (beforeSMOTE <- prop.table(table(train$target)))
##
## 0 1
## 0.5 0.5
##
## 0 1
## 0.95322377 0.04677623
train$target <- as.numeric(train$target); train.SMOTE$target <- as.numeric(train.SMOTE$target) #revert
#We then train using the SMOTE 'd training set and predict using the same testing set as used before on the non- SMOTE 'd training set to ensure we're comparing apples-to-apples:
bagModel.SMOTE <- train(target ~ ., data = train.SMOTE, method = "treebag", trControl = ctrl)
Predicts.SMOTE <- predict(bagModel.SMOTE$finalModel, test[,predictors])
auc.SMOTE <- roc(test$target, Predicts.SMOTE)
auc.SMOTE[9]
## $auc
## Area under the curve: 0.9933