arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2202.11014math.OC
收藏

用Hamilton-Jacobi偏微分方程的演化方法求非凸问题的整体解

Global Solutions to Nonconvex Problems by Evolution of Hamilton-Jacobi PDEs

Howard Heaton, Samy Wu Fung, Stanley Osher

摘要:

计算任务通常可以被视为最优化问题。真实世界场景的目标函数通常是非凸的和/或不可微的。解决这些问题的最先进的方法通常只能保证收敛到局部最小值。本文提出了一种基于Hamilton-Jacobi的Moreau自适应下降算法(HJ-MAD),该算法在假设目标函数连续的情况下,保证了算法的全局最小收敛性。其核心思想是计算梯度的莫罗包络的目标(这是“分段凸”)与自适应平滑参数。Moreau包络线\rev{(\ie proximal operators)}的分量通过粘性Hamilton-Jacobi方程的Hopf-Lax公式近似。我们的数值例子说明了整体收敛。

英文摘要:

Computing tasks may often be posed as optimization problems. The objective functions for real-world scenarios are often nonconvex and/or nondifferentiable. State-of-the-art methods for solving these problems typically only guarantee convergence to local minima. This work presents Hamilton-Jacobi-based Moreau Adaptive Descent (HJ-MAD), a zero-order algorithm with guaranteed convergence to global minima, assuming continuity of the objective function. The core idea is to compute gradients of the Moreau envelope of the objective (which is "piece-wise convex") with adaptive smoothing parameters. Gradients of the Moreau envelope \rev{(\ie proximal operators)} are approximated via the Hopf-Lax formula for the viscous Hamilton-Jacobi equation. Our numerical examples illustrate global convergence.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑