You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I built this model for my btyd library way back in 2022. It sacrifices performance (both in speed and posterior estimation) so that non-repeat customers can have an alive probability < 1. I don't consider this trade-off worthwhile given the effort required to implement this model, and won't be working on it. That being said, if anyone else wishes to do so, they're more than welcome.
Very similar to #67, except it allows for dropout before the first purchase: https://github.com/CamDavidsonPilon/lifetimes/blob/master/lifetimes/fitters/modified_beta_geo_fitter.py
The text was updated successfully, but these errors were encountered: