A generative flow for conditional sampling via optimal transport
This work addresses conditional sampling, a fundamental task in Bayesian inference, with an incremental extension of existing data-driven methods.
The paper tackles the problem of sampling conditional distributions for Bayesian inference and density estimation by proposing a non-parametric generative model using block-triangular transport maps derived from optimal transport, and demonstrates it on a 2D example and a nonlinear ODE parameter inference problem.
Sampling conditional distributions is a fundamental task for Bayesian inference and density estimation. Generative models, such as normalizing flows and generative adversarial networks, characterize conditional distributions by learning a transport map that pushes forward a simple reference (e.g., a standard Gaussian) to a target distribution. While these approaches successfully describe many non-Gaussian problems, their performance is often limited by parametric bias and the reliability of gradient-based (adversarial) optimizers to learn these transformations. This work proposes a non-parametric generative model that iteratively maps reference samples to the target. The model uses block-triangular transport maps, whose components are shown to characterize conditionals of the target distribution. These maps arise from solving an optimal transport problem with a weighted $L^2$ cost function, thereby extending the data-driven approach in [Trigila and Tabak, 2016] for conditional sampling. The proposed approach is demonstrated on a two dimensional example and on a parameter inference problem involving nonlinear ODEs.