Reproduce “Stochastic set-valued optimization and its application to robust learning” 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
Tommaso Giovannelli, Jingfu Tan, Luis Nunes Vicente
The authors develop a method for binary classification that treats lower- and upper-tail loss assessments as separate objectives and selects a compromise on their Pareto trade-off. They extend stochastic set-valued optimization to two- and four-objective formulations and characterize subquantiles. On Adult, logistic regression showed lower test-loss variability than empirical risk minimization across tested positive-class fractions under the replication protocol.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.