Reproduce “Collaborative Filtering with Side Information: a Gaussian Process Perspective” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Hyunjik Kim, Xiaoyu Lu, Seth Flaxman, Yee Whye Teh
Tucker Gaussian Process (TGP) predicts missing user–item ratings by combining Gaussian-process kernels for user and item similarities, including side information, with low-rank Tucker factorization. This approximates a direct Gaussian-process model. On MovieLens 100K, side information lowers test RMSE from 0.9270 ± 0.0097 to 0.8995 ± 0.0062; this five-split MAP result does not establish statistical significance or general superiority.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.