Skip to yearly menu bar Skip to main content


Poster

Multi-objective Optimization via Wasserstein-Fisher-Rao Gradient Flow

Yinuo Ren · Tesi Xiao · Tanmay Gangwani · Anshuka Rangi · Holakou Rahmanian · Lexing Ying · Subhajit Sanyal

MR1 & MR2 - Number 131

Abstract:

Multi-objective optimization (MOO) aims to optimize multiple, possibly conflicting objectives with widespread applications. We introduce a novel interacting particle method for MOO inspired by molecular dynamics simulations. Our approach combines overdamped Langevin and birth-death dynamics, incorporating a ``dominance potential'' to steer particles toward global Pareto optimality. In contrast to previous methods, our method is able to relocate dominated particles, making it particularly adept at managing Pareto fronts of complicated geometries. Our method is also theoretically grounded as a Wasserstein-Fisher-Rao gradient flow with convergence guarantees. Extensive experiments confirm that our approach outperforms state-of-the-art methods on challenging synthetic and real-world datasets.

Chat is not available.