Model Evaluation.md

Welcome to GitHub Pages

Model Evaluation using 6 different Machine Learning Algorithms in R, for Breast Cancer dataset

1. Random Forest

Load the package

library (randomForest)

Read file_

all_data <- read.csv (file='D:/model_evaluation/all_data.csv')

Run randomForest

all_data.rf <- randomForest (V24 ~. , data=all_data, ntree=500, mtry=5, keep.forest=FALSE, importance=TRUE)

Print result of randomForest (OOB error and confusion matrix)

print (all_data.rf)

2. Decision Tree

Load the packages

library (caret) library (rpart)

Split data into testing and training

intrain <- createDataPartition (y = all_data $V24, p= 0.7, list = FALSE) training <- all_data [intrain,] testing <- all_data [-intrain,]

Train the model

trctrl <- trainControl (method = "repeatedcv", number = 10, repeats = 3) dtree_fit <- train (V24 ~. , data = training, method = "rpart", parms = list (split = "information"), trControl=trctrl, tuneLength = 10)

Prediction using testing data

predict (dtree_fit, newdata = testing) test_pred <- predict (dtree_fit, newdata = testing)

print result of randomForest (Accuracy and confusion matrix)

confusionMatrix (test_pred, testing$V24)

3. Support Vector Machine (SVM)

Load the packages

library(caret)

Assign the categorical variable as a factor in both training and testing sets (use training and testing sets from Decision Tree)

training[["V24"]] = factor(training[["V24"]]) testing[["V24"]] = factor(testing[["V24"]])

Train the svm model

trctrl <- trainControl(method = "repeatedcv", number = 10, repeats = 3) set.seed(3233) svm_Linear <- train(V24 ~., data = training, method = "svmLinear",trControl=trctrl, preProcess = c("center", "scale"), tuneLength = 10) svm_Linear

Test set prediction

test_pred <- predict(svm_Linear, newdata = testing) test_pred

Compute accuracy using confusion matrix

confusionMatrix(test_pred, testing$V24)

4. Logistic Regression

Training the dataset with lm

lmMod <- lm(V24 ~ ., data=training) #build the model distPred <- predict(lmMod, testing) #predict survival summary (lmMod)

Test set prediction

actuals_preds <- data.frame(cbind(actuals=testing$V24, predicteds=distPred)) #make actuals_predicteds dataframe correlation_accuracy <- cor(actuals_preds) head(actuals_preds)

5. Neural Networks

Load packages

library(caret) library(neuralnet)

Fit neural network

set.seed(2) NN = neuralnet(V24 ~ V1+V2+V3+V4+V5+V6+V7+V8+V9+V10+V11+V12+V13+V14+V15+V16+V17+V18+V19+V20+V21+V22+V23, training, hidden = 3 , linear.output = T)

Prediction using neural network

predict_testNN = compute(NN, testing[,c(1:23)]) predict_testNN = (predict_testNN$net.result * (max(all_data$V24) - >min(all_data$V24))) + min(all_data$V24) plot(testing$V24, predict_testNN, col='blue', pch=16, ylab = "predicted rating NN", xlab = "real rating") abline(0,1)

Calculate Root Mean Square Error (RMSE)

RMSE.NN = (sum((testing$V24 - predict_testNN)^2) / nrow(testing)) ^ 0.5

6. Extreme Gradient Boosting

Load packages

library(xgboost) library(caret)

Train with xgboost

folds = createFolds(training$V24, k=10) cv = lapply (folds, function(x){ training_fold= training[-x,] testing_fold= testing[x,] classifier = xgboost(data=as.matrix(training[-24]), label=training$V24, nrounds=10 ) y_pred = predict(classifier, newdata=as.matrix(testing_fold[24])) y_pred =(y_pred >= 0.5) cm = table(testing_fold[,24], y_pred) accuracy = (cm[1,1] + cm[2,2]) / (cm[1,1] + cm[2,2] + cm[1,2] + cm[2,1]) return (accuracy) }) accuracy = mean(as.numeric(cv))