Model Evaluation: ROC-AUC Curves

Machine Learning can be really confusing at times but don’t worry, you are at the right place! The terms might be intimidating but once you get to know the details, it’s really easy!

WHAT IS ROC?!

A ROC curve (Receiver Operating Characteristic Curve) is a graph showing the performance of a classification model at all classification thresholds

Before going into detail, let’s get to know the history behind the curves.

HISTORY OF ROC:

The origin of ROC curve goes way back to World War II, it was originally used for the analysis of radar signals. The United States Army tried to measure the ability of their radar receiver to correctly identify the Japanese Aircraft.

ROC CURVE -CLASSIFICATION MADE EASIER:

A ROC Curve is plotted against True Positive Rate vs False Positive Rate

CONFUSION MATRIX

TPR and FPR can be calculated by the following formula:

The larger the TPR, the better the model.

“False hopes are more dangerous than fears.”–J.R.R. Tolkein

AUC CURVE:

AUC stands for “Area under the ROC Curve.”

AUC measures the entire two-dimensional area underneath the entire ROC curve.

Thus these curves help us analyze the performance metrics of the model we developed.

That’s it for today folks!