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 2512.13993math.NAcs.NAmath.OC
收藏

多尺度方法用于离散化连续函数的优化

Multiscale Methods for Discretized Continuous Optimization: Convergence and Cost Analysis

University of British Columbia(不列颠哥伦比亚大学)

Nicholas J. E. Richardson, Noah Marusenko, Michael P. Friedlander

AI总结:

本文提出一种多尺度优化方法,通过粗网格预热启动梯度下降,实现更高效的连续函数优化,适用于概率密度估计等应用。

摘要:

本文开发了一种多尺度优化框架,用于解决在Lipschitz连续函数空间上的问题。该方法通过粗网格离散化后进行线性插值,以预热启动梯度下降算法,逐步在更细的网格上进行优化。分析了贪婪和懒惰变种,并推导出收敛保证,证明多尺度方法在较低计算成本下能够实现更紧的误差界,优于单尺度优化。分析扩展到任何具有固定收敛率的基算法。约束修改技术在不同尺度上保持可行性。在概率密度估计问题上的数值实验,包括地质数据,展示了速度提升一个数量级或更好的效果。

英文摘要:

Discretized versions of optimization problems over continuous arguments are routinely solved at a single fine resolution, incurring a per-iteration cost that grows, often superlinearly, with the number of grid points. This paper analyzes a multiscale method that instead solves a hierarchy of increasingly fine dyadic discretizations. Linear interpolation of each coarse solution warm starts the next finer scale using any q-linearly convergent update rule as the inner solver. Each coarse problem is a consistent discretization of the continuous problem. Structural properties such as convexity and smoothness are preserved. For problems with Lipschitz-continuous solutions, two variants of the method converge to the fine-scale solution with explicit error bounds. The fine-scale solution in turn approximates the continuous solution once the grid is sufficiently fine, with quantified constants. The total cost to reach a fixed accuracy is provably lower than that of single-scale optimization whenever the cost of one update grows at least linearly in the problem size. Numerical experiments on probability density demixing problems, including geological survey data, show four- to sevenfold speedups while using a fraction of the memory.

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。

↑