Reproduce “Normalized Cuts and Data Clustering via Matrix Fractional Programming” 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
Yannan Chen, Beichen Huang, Licheng Zhao, Chris Ding, Kaiming Shen
FPC directly optimizes multiclass normalized cut (NCut) clustering by transforming ratios into per-point cluster scores and assigning each point to its highest-scoring cluster. On Sensorless, it reports an NCut objective of 1.151889, 5.14% below spectral clustering’s 1.214324. It is best or tied-best on eleven of twelve datasets, but lower NCut does not establish better labels or global optimality.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.