Neural Predictor-Corrector: Solving Homotopy Problems with Reinforcement Learning

Jiayao Mai, Bangyan Liao, Zhenjun Zhao +6

The Homotopy paradigm, a general principle for solving challenging problems, appears across diverse domains such as robust optimization, global optimization, polynomial root-findin…