Accuracy Plots

• Naive Bayes ROC Curve

• J48 Decision Tree ROC Curve

• JRip ROC Curve

• SVM ROC Curve

1. The  Figure below represents the overall ROC curve for each of the classifiers applied to our dataset .

• • As we can see, J48 Decision Tree classifier and JRip give nearly perfect classification , according to the cross validation experiments on the training dataset.

  • Their ROC curves are  very close to the upper left corner which proves that the overall accuracy is high.

  • On the other hand, Naive Bayes and SVM performed significantly poorly in comparison to the above two although still very well in absolute terms .  

  • The ROC curves for Naive Bayes and SVM classifiers also show the highest deviation from the upper left corner which significantly is poor in comparison to J48 Decision Tree ROC Curve .