Reproduce “Dimensionality Reduction and (Bucket) Ranking: a Mass Transportation Approach” 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
Mastane Achab, Anna Korba, Stephan Clémençon
The authors introduce bucket distributions to summarize ranking data by fixing the order between groups of items while retaining observed variation within each group. Their Kendall’s tau distortion equals the expected number of cross-group order violations. Across sushi and car preference datasets, lower-dimensional representations sometimes retain low empirical distortion, but this is not held-out validation or evidence of downstream gains.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.