New Method Aims for Robust Distributional Learning from Contaminated Data
A new arXiv paper proposes a robust-by-design approach to distributional learning that handles datasets assembled via retrieval, scraping, automated filtering, and post-hoc curation, which are prone to contamination.
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-11-02 05:00 · arXiv stat.ML
Robust-by-Design Distributional Learning from Contaminated Samples