Closed alexliap closed 10 months ago
Hi @alexliap.
I developed an ML algorithm (a simple decision tree) capable of predicting the monotonicity maximizing the Information Value. The features of the model are designed to characterize a prebinning solution. The decision tree is simple, so to make it fast, it is implemented as a set of rules. The paper explains in detail the training process and the resulting prediction metrics.
Could you please explain how you predict the woe monotonicity for a variable? I tried to read the relevant paper but didn't quite understand it. Up to now in my company we have been figuring it out by binning the variable manually. It would be very useful to understand how its done automatically and not just take it as a black box.