Generalization Bound and Learning Methods for Data-Driven Projections in Linear Programming

Shinsaku Sakaue, The University of Tokyo and RIKEN AIP, Tokyo, Japan, sakaue@mist.i.u-tokyo.ac.jp, "3026 Taihei Oki, Hokkaido University, Hokkaido, Japan, oki@icredd.hokudai.ac.jp

Feb-8-2026, 13:24:42 GMT–Neural Information Processing Systems 

Given training data ofn-dimensional LPs, we learn ann k projection matrix withn>k.