Reproduce “Fitting Spectral Decay with the $k$-Support Norm” 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

Andrew M. McDonald, Massimiliano Pontil, Dimitris Stamos

The authors generalize the spectral k-support norm for matrix completion using tunable p to change how singular values are penalized. At p=2, it recovers the original norm; p=infinity favors a flatter spectrum. On MovieLens 100k, it reports NMAE of 0.1921, versus 0.1990 for spectral k-support and 0.2017 for trace norm. Parameters were selected by validation, and results do not guarantee improvement elsewhere.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.