think of an intro (something about how you haven’t done serious ML in a while and you decided to blitz through some GAN papers - maybe talk about how you want to colour manga using GANS (link to the paper) and how the paper didn’t make any sense at first)

Before talking about GANS i think it’d be cool to talk about how computer scientists frame the problem of generating (data?)

the initial model for candidate selection was WRMF(weight regularized matrix factorization)
from the name its pretty easy to assume what it’s supposed to be doing

basically matrix factorization as used in regular collaborative filtering problems

where the userxitem ie preferences/ratings or whatever floats your boat is factorized into user and item embeddings that can be used for subsequent predictions

I think the regularization employed here is just for more efficiently learning (something i think is pretty relevant in this case since we’ve got like 20k…..) the creators of v16 link to a paper that more or less describes their specific approach to WRMF (Collaborative Filtering for Immediate Feedback Datasets)