17 March 2023 2 9K Report

Hello !

In LDA there are different ways to assess which is the 'best' number of topics. I would like to evaluate models with different k1 and k2 in Pachinko topic modeling in Python (package tomotopy currently, but I could switch to another package if it's more practical for this). Would someone know how this can be done?

Best wishes,

Eric

More Eric Mayor's questions See All
Similar questions and discussions