This notebook accompanies the paper "Variational autoencoders for collaborative filtering" by Dawen Liang, Rahul G. Krishnan, Matthew D. Hoffman, and Tony Jebara, in The Web Conference (aka WWW) 2018.
In this notebook, we show a complete self-contained example of training a variational autoencoder (as well as a denoising autoencoder) with multinomial likelihood (described in the paper) on the public Movielens-20M dataset, including both data preprocessing and model training.