Reflected anchored Langevin algorithm samples nonsmooth constrained distributions
Reflected Anchored Langevin Algorithms
Machine Learning
Summary
Sampling from complicated data distributions that have hard boundaries and sharp corners is a common problem in machine learning, but existing methods need smooth shapes to work well. The authors introduce a new approach called reflected anchored Langevin dynamics that can handle sampling from nonsmooth and constrained spaces by combining a smooth reference shape with a special scaling in the process's movement and noise. They develop a practical algorithm from this idea and prove it gets closer to the target distribution effectively. Their experiments confirm the theory and show the method can handle cases traditional methods struggle with.
What this means in practice
- •For machine learning engineers: Handle sampling from constrained probability models with nonsmooth log densities to better train or evaluate models with hard boundaries.
- •For statistical modelers: Generate samples from complex constrained distributions where traditional gradient-based methods fail due to non-differentiability.
Authors
Changwei Tu, Xiaoyu Wang, Yingli Wang, Xicheng Zhang, Lingjiong Zhu
Abstract
First order Langevin algorithms for constrained sampling in machine learning, such as projected Langevin Monte Carlo which are based on discretizations of reflected Langevin dynamics, require differentiable log densities that limits their applicability. This paper introduces reflected anchored Langevin dynamics (RALD), a reflected diffusion that converges to non-differentiable targets on constrained domains. The method uses a smooth anchored reference potential and multiplies the drift and noise covariance of its reflected Langevin dynamics by the same state dependent scaling factor. Its Euler-Maruyama discretization with projection gives reflected anchored Langevin Monte Carlo (RALMC) algorithm. We prove explicit convergence bounds and iteration complexity for RALMC in the 2-Wasserstein distance to the target distribution. Numerical experiments are provided to illustrate the theoretical predictions and the empirical performance of the method.