Thanks for the pointers, Sylvian. I am aware of some techniques to find optimal # of bins and their limits so that least amount of information is lost in discretization and the model performs well. Applicable when one plans to use the discretized features instead of the continuous ones in the model. Will go through the links that you’ve provided.

Not completely sure if this will help in this scenario where we are binning only for exploration purpose. It could avoid creating too many or too fewer bins. Could also give more accurate results for model monitoring/ feature drift.

Thanks again!