SNE_lab 0.1.5

Structured Neural Embedding model for research

This module aims to implement Structured Neural Embedding, a novel multi-task multi-label predictor, upon NumPy. It runs on MovieLens100K, MovieLens1M, Youtube, and Ego-net (Facebook)’s network datasets. The original purpose of it was to serve as one of the baseline models in the paper Deep Energy Factorization Model for Demographic Prediction (Chih-Te Lai, Po-Kai Chang, et al). Give SNE-lab a try if either you need a handy implementation or you want a quick overview on multi-task multi-label prediction!