Reproduce “More Powerful Selective Kernel Tests for Feature Selection” 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
Jen Ning Lim, Makoto Yamada, Wittawat Jitkrittum, Yoshikazu Terada, Shigeyuki Matsui, Hidetoshi Shimodaira
The authors develop feature-specific selective inference for kernel tests, conditioning each selected feature’s test only on its own selection rather than the full selected set. They use selective multiscale bootstrap and add an incomplete HSIC estimator for dependence testing. The methods are more powerful in most reported scenarios; in the pulsar benchmark, TPR is 0.993 versus 0.746, with FPRs of 0.056 and 0.063.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.