Abstract:We present the fundamentals of a measure transport approach to sampling. The idea is to construct a deterministic coupling---i.e., a transport map---between a complex "target" probability measure of interest and a simpler reference measure. Given a transport map, one can generate arbitrarily many independent and unweighted samples from the target simply by pushing forward reference samples through the map. We consider two different and complementary scenarios: first, when only evaluations of the unnormalized target density are available, and second, when the target distribution is known only through a finite collection of samples. We show that in both settings the desired transports can be characterized as the solutions of variational problems. We then address practical issues associated with the optimization--based construction of transports: choosing finite-dimensional parameterizations of the map, enforcing monotonicity, quantifying the error of approximate transports, and refining approximate transports by enriching the corresponding approximation spaces. Approximate transports can also be used to "Gaussianize" complex distributions and thus precondition conventional asymptotically exact sampling schemes. We place the measure transport approach in broader context, describing connections with other optimization--based samplers, with inference and density estimation schemes using optimal transport, and with alternative transformation--based approaches to simulation. We also sketch current work aimed at the construction of transport maps in high dimensions, exploiting essential features of the target distribution (e.g., conditional independence, low-rank structure). The approaches and algorithms presented here have direct applications to Bayesian computation and to broader problems of stochastic simulation.

The Transport Map Computed by Iterated Function System

Constrained Approximate Optimal Transport Maps

Functional optimal transport: map estimation and domain adaptation for functional data

An Approximation Theory Framework for Measure-Transport Sampling Algorithms

Optimal transportation and stationary measures for Iterated Function Systems

Optimal transport map estimation in general function spaces

Optimal Transport of Linear Systems over Equilibrium Measures

On optimal transport of matrix-valued measures

Statistical Convergence Rates of Optimal Transport Map Estimation between General Distributions

Approximating Optimal Transport via Low-rank and Sparse Factorization

A Consistent Extension of Discrete Optimal Transport Maps for Machine Learning Applications

An introduction to sampling via measure transport

Learning Cost Functions for Optimal Transport

Fast and Accurate Approximations of the Optimal Transport in Semi-Discrete and Discrete Settings

Using Linearized Optimal Transport to Predict the Evolution of Stochastic Particle Systems

Entropic Approximation of -Optimal Transport Problems

Expected Sliced Transport Plans

Discrete Probabilistic Inverse Optimal Transport

Deep composition of tensor-trains using squared inverse Rosenblatt transports

A Distributed Framework for the Construction of Transport Maps

Semi-Discrete approximation of Optimal Mass Transport