Penalized Langevin and Hamiltonian Monte Carlo Algorithms for Constrained Sampling

11/29/2022
by   Mert Gurbuzbalaban, et al.
0

We consider the constrained sampling problem where the goal is to sample from a distribution π(x)∝ e^-f(x) and x is constrained on a convex body 𝒞⊂ℝ^d. Motivated by penalty methods from optimization, we propose penalized Langevin Dynamics (PLD) and penalized Hamiltonian Monte Carlo (PHMC) that convert the constrained sampling problem into an unconstrained one by introducing a penalty function for constraint violations. When f is smooth and the gradient is available, we show 𝒪̃(d/ε^10) iteration complexity for PLD to sample the target up to an ε-error where the error is measured in terms of the total variation distance and 𝒪̃(·) hides some logarithmic factors. For PHMC, we improve this result to 𝒪̃(√(d)/ε^7) when the Hessian of f is Lipschitz and the boundary of 𝒞 is sufficiently smooth. To our knowledge, these are the first convergence rate results for Hamiltonian Monte Carlo methods in the constrained sampling setting that can handle non-convex f and can provide guarantees with the best dimension dependency among existing methods with deterministic gradients. We then consider the setting where unbiased stochastic gradients are available. We propose PSGLD and PSGHMC that can handle stochastic gradients without Metropolis-Hasting correction steps. When f is strongly convex and smooth, we obtain an iteration complexity of 𝒪̃(d/ε^18) and 𝒪̃(d√(d)/ε^39) respectively in the 2-Wasserstein distance. For the more general case, when f is smooth and non-convex, we also provide finite-time performance bounds and iteration complexity results. Finally, we test our algorithms on Bayesian LASSO regression and Bayesian constrained deep learning problems.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
08/27/2021

An Introduction to Hamiltonian Monte Carlo Method for Sampling

The goal of this article is to introduce the Hamiltonian Monte Carlo (HM...
research
06/24/2020

Penalized Langevin dynamics with vanishing penalty for smooth and log-concave targets

We study the problem of sampling from a probability distribution on ℝ^p ...
research
10/13/2022

Condition-number-independent Convergence Rate of Riemannian Hamiltonian Monte Carlo with Numerical Integrators

We study the convergence rate of discretized Riemannian Hamiltonian Mont...
research
05/07/2019

Optimal Convergence Rate of Hamiltonian Monte Carlo for Strongly Logconcave Distributions

We study Hamiltonian Monte Carlo (HMC) for sampling from a strongly logc...
research
12/07/2020

Adaptive Sequential SAA for Solving Two-stage Stochastic Linear Programs

We present adaptive sequential SAA (sample average approximation) algori...
research
01/31/2023

Non-convex sampling for a mixture of locally smooth potentials

The purpose of this paper is to examine the sampling problem through Eul...
research
02/09/2021

A New Framework for Variance-Reduced Hamiltonian Monte Carlo

We propose a new framework of variance-reduced Hamiltonian Monte Carlo (...

Please sign up or login with your details

Forgot password? Click here to reset