Optimization method improves search on unknown shapes from sample points
Riemannian Optimization on Unknown Manifolds via Stochastic Proximal Refinement
Machine Learning
Summary
Sometimes people want to find the best solution that fits certain rules defined by a complicated shape, but they only have a few examples that follow the rules instead of a clear description of the shape. The authors designed a way called PRISM that uses these example points to take smart steps toward better solutions without needing to know the shape’s details. Their method keeps solutions valid and gradually improves the objective being optimized. They tested their idea on made-up examples and real control and optimization tasks, showing it works well in practice.
What this means in practice
- •For building control engineers: Optimize building control settings using limited sensor data without explicit system models, improving energy efficiency and comfort.
- •For industrial optimization teams: Improve solutions for complex engineering problems when only sample-based feasible data is available instead of explicit constraints.
Authors
Ruixiang Wu, Jiahao Ai, Tinko Sebastian Bartels, Xuanzhuo Liu, Yifei Wu, Yu Mao, Tongxin Li, Costas Courcoubetis
Abstract
We study Riemannian optimization on an unknown manifold accessed only through feasible samples. The central challenge is to turn feasible samples into optimization steps, i.e., the samples are drawn from the unknown feasible manifold, but do not directly provide the tangent spaces, projections, or retractions required by classic Riemannian methods. Moreover, regions with low objective values may be sparsely represented in the data, so directly sampling from a distribution fitted to these observations can require many draws to find better feasible solutions. We introduce Proximal Refinement on Implicit Smooth Manifolds (PRISM), which turns feasible-sample access into an iterative optimization procedure by repeatedly sampling an objective-aware proximal distribution conditioned on the current iterate. Theoretically, we show that PRISM recovers the effect of Riemannian-gradient refinement, leading to provable objective descent, feasibility (manifold adherence), and convergence toward first-order stationarity. We demonstrate the effectiveness of our approach through synthetic experiments illustrating its geometric properties and applications to data-driven building control and diverse real world optimization tasks, highlighting its potential as a general framework for data-driven optimization across scientific and engineering domains.