In this project, we conduct research on collaborative filtering approach with latent factor decomposition, a matrix factorization approach to decompose user-item interaction information to recommend users’ most relevant items, with Alternative Least Square method. We utilize PySpark (version 2.4.0) and implement baseline model on HPC environment. We further implement model on single-machine to compare its performance from HPC settings, and conduct various visualization to detect the hidden correlations between latent factors and other features.
If you have any questions, you can reach out to me hg1153@nyu.edu