Hey data scientists of Reddit, would like to seek some advice from the experts here.

I am trying to implement an assortment optimization problem, where you have to pick a subset of good from the master set which maximises the objective (i.e. revenue/profit). Eventually, I would like to work towards a tractable SKU-level sale forecast function (choice model) that accounts for cannibalization, substitution and seasonality. As I understand it, this function would have to be fed into an optimisation algorithm (i.e. linear integer programming).

From my readings, I gather that Multinomial Logit Model (MNL) should be the baseline/ starting point. However, I can’t seem to grasp how I could extend MMNL to account for cannibalization, substitution and seasonality.

Can any kind souls point me to a good place to start on optimization in the field of operations research for assortment optimization?

​

tldr: Where can I find resources on assortment optimization with cannibalization, substitution and seasonality effects?