Pareto Optimization with Robust Evaluation for Noisy Subset Selection

Yiheng Xu, Danxuan Liu, Bin Zhang, Weiyong Yang, Chao Qian · 2025

Read Pareto Optimization with Robust Evaluation for Noisy Subset Selection by Yiheng Xu, Danxuan Liu, Bin Zhang, Weiyong Yang, Chao Qian (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-06813.txt.

Description

Yiheng Xu and four co-authors pick subsets under noise with a method that judges each pick by its worst credible showing. Choosing a small set with the best total effect drives tasks from spreading influence to sparse modeling. In practice the score of any set is measured with error, and greedy or older evolutionary methods either wobble under that noise or burn too much computing. The team's method, PORE, balances two goals at once: a robust score that discounts lucky noise, and a small set size.

That balance lets it find well-shaped solutions while spending its budget wisely, fixing a known waste in a predecessor method. On real data for influence spread and sparse regression, the method clearly beat the greedy rule and both earlier evolutionary rivals. Follow-up tests that removed parts of the design confirmed the robust score does the heavy lifting.

Catalog record

CREATORYiheng Xu, Danxuan Liu, Bin Zhang, Weiyong Yang, Chao Qian
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE39 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-06813

Browse the full library · About the Tower of Babel