Eugene Yan • 1/13/2020

Beating the Baseline Recommender with Graph & NLP in Pytorch

This technical article details an approach to beat a baseline matrix factorization recommender system. It combines Natural Language Processing (word2vec) and graph algorithms (DeepWalk, random walks) to learn product embeddings from a graph of product relationships, implemented in PyTorch and NetworkX.

0 comments

#Machine Learning #Natural Language Processing #Pytorch