Reproduce “Discovery of fully efficient fault indicators along a data-based diagnosis process” 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
Igor Bezmaternykh, Louise Travé-Massuyès, Elodie Chanthery
DT4X+ changes how each decision tree split is learned, retaining a readable multivariate expression that separates two target classes while penalizing splits that scatter non-target classes across branches. It adds non-target classes to training and the split loss. On steam superheater 4.2 cyber-attack-detection, accuracy rises from 0.98968 with DT4X to 0.99479 with DT4X+, but gains are not uniform across benchmarks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.