Sampling from Log-Concave Distributions over Polytopes via a Soft-Threshold Dikin Walk
This work addresses a computational bottleneck in machine learning applications like Bayesian inference and privacy-preserving learning, offering an incremental improvement over prior methods for specific parameter settings.
The paper tackles the problem of sampling from log-concave distributions over polytopes, which is important for Bayesian inference and differentially private learning, by introducing a soft-threshold Dikin walk algorithm that improves running time to O((md + d L^2 R^2) × md^{ω-1}) log(w/δ)) operations for sampling within error δ from a warm start.
Given a Lipschitz or smooth convex function $\, f:K \to \mathbb{R}$ for a bounded polytope $K \subseteq \mathbb{R}^d$ defined by $m$ inequalities, we consider the problem of sampling from the log-concave distribution $π(θ) \propto e^{-f(θ)}$ constrained to $K$. Interest in this problem derives from its applications to Bayesian inference and differentially private learning. Our main result is a generalization of the Dikin walk Markov chain to this setting that requires at most $O((md + d L^2 R^2) \times md^{ω-1}) \log(\frac{w}δ))$ arithmetic operations to sample from $π$ within error $δ>0$ in the total variation distance from a $w$-warm start. Here $L$ is the Lipschitz-constant of $f$, $K$ is contained in a ball of radius $R$ and contains a ball of smaller radius $r$, and $ω$ is the matrix-multiplication constant. Our algorithm improves on the running time of prior works for a range of parameter settings important for the aforementioned learning applications. Technically, we depart from previous Dikin walks by adding a "soft-threshold" regularizer derived from the Lipschitz or smoothness properties of $f$ to the log-barrier function for $K$ that allows our version of the Dikin walk to propose updates that have a high Metropolis acceptance ratio for $f$, while at the same time remaining inside the polytope $K$.