Sampling in Combinatorial Spaces with SurVAE Flow Augmented MCMC
arXiv:2102.02374
Abstract
Hybrid Monte Carlo is a powerful Markov Chain Monte Carlo method for sampling from complex continuous distributions. However, a major limitation of HMC is its inability to be applied to discrete domains due to the lack of gradient signal. In this work, we introduce a new approach based on augmenting Monte Carlo methods with SurVAE Flows to sample from discrete distributions using a combination of neural transport methods like normalizing flows and variational dequantization, and the Metropolis-Hastings rule. Our method first learns a continuous embedding of the discrete space using a surjective map and subsequently learns a bijective transformation from the continuous space to an approximately Gaussian distributed latent variable. Sampling proceeds by simulating MCMC chains in the latent space and mapping these samples to the target discrete space via the learned transformations. We demonstrate the efficacy of our algorithm on a range of examples from statistics, computational physics and machine learning, and observe improvements compared to alternative algorithms.
arXiv commentAccepted at AISTATS 2021; added experiments with longer MCMC chains
References in corpus (11)
- PyTorch: An Imperative Style, High-Performance Deep Learning Library
- MCMC using Hamiltonian dynamics
- Variational Inference with Normalizing Flows
- Flow-based generative models for Markov chain Monte Carlo in lattice field theory
- Flow++: Improving Flow-Based Generative Models with Variational Dequantization and Architecture Design
- An introduction to sampling via measure transport
- RNADE: The real-valued neural autoregressive density-estimator
- NeuTra-lizing Bad Geometry in Hamiltonian Monte Carlo Using Neural Transport
- Discontinuous Hamiltonian Monte Carlo for discrete parameters and discontinuous likelihoods
- Auxiliary-variable Exact Hamiltonian Monte Carlo Samplers for Binary Distributions
- Accelerated Sampling on Discrete Spaces with Non-Reversible Markov Processes