I trained a binary classifier using scikit-learn and plotted some ROC curves to evaluate it. The problem is I have to present it to someone else and ROC curves are not so intuitive.
I found a nice graph showing the probability density distributions (https://kennis-research.shinyapps.io/ROC-Curves/) and believe it is a very nice tool to easily present the results and explain the concepts of classification to my audience. I tried to find this type of representation on scikit-learn, and tried to create it based on the results of the calls classifier.fit.predict_proba() and roc_curve methods, but no success so far.
Do you know any library that has this type of representation, or could give me a hint on how to do it using the classifier/fit/predict data?
Thanks, Colis
