gist_2

#FORESTS --------------------------
rf=randomForest(diagnosis~.,data=mydata.train,ntree=250, mtry = 8)
predrf=predict(rf,newdata=mydata.valid)
forest_accuracy = mean(predrf==mydata.valid$diagnosis)

# SVM ------------------------------
mysvm = svm(diagnosis~., data = mydata.train, kernel="polynomial", cost=5, degree=3)
pred_svm_optimal = predict(mysvm, mydata.valid)
accuracy_svm = mean(pred_svm_optimal==mydata.valid$diagnosis)  

# LOGISTIC REGRESSION -----------------
logistic = glm(mydata.train$diagnosis~., data = mydata.train, family = binomial)
pred = round(predict(logistic, type = "response", newdata=mydata.valid))
accuracy_logistic = mean(pred==mydata.valid$diagnosis)