Hierarchical Implicit Models and Likelihood-Free Variational Inference

TL;DR

This paper introduces hierarchical implicit models (HIMs) that combine implicit densities with hierarchical Bayesian structures, and develops likelihood-free variational inference (LFVI) for scalable inference in these models, demonstrated across diverse applications.

Contribution

The paper proposes HIMs for complex hierarchical modeling with implicit densities and introduces LFVI, a scalable inference algorithm for such models, enabling broader application.

Findings

Successful application to ecological predator-prey simulation

Effective Bayesian GAN for discrete data

Deep implicit model for text generation

Abstract

Implicit probabilistic models are a flexible class of models defined by a simulation process for data. They form the basis for theories which encompass our understanding of the physical world. Despite this fundamental nature, the use of implicit models remains limited due to challenges in specifying complex latent structure in them, and in performing inferences in such models with large data sets. In this paper, we first introduce hierarchical implicit models (HIMs). HIMs combine the idea of implicit densities with hierarchical Bayesian modeling, thereby defining models via simulators of data with rich hidden structure. Next, we develop likelihood-free variational inference (LFVI), a scalable variational inference algorithm for HIMs. Key to LFVI is specifying a variational family that is also implicit. This matches the model's flexibility and allows for accurate approximation of the posterior. We demonstrate diverse applications: a large-scale physical simulator for predator-prey populations in ecology; a Bayesian generative adversarial network for discrete data; and a deep implicit model for text generation.