Importance Weighted Generative Networks

TL;DR

This paper introduces importance weighted methods for deep generative networks that enable training with biased or indirectly accessible target distributions, providing theoretical guarantees and strong empirical results.

Contribution

It proposes importance weighting techniques to estimate losses for target distributions without direct access, enhancing the robustness of generative models.

Findings

The methods achieve accurate loss estimation under sample bias.

The approach offers theoretical guarantees for convergence.

Empirical results demonstrate improved generative performance.

Abstract

Deep generative networks can simulate from a complex target distribution, by minimizing a loss with respect to samples from that distribution. However, often we do not have direct access to our target distribution - our data may be subject to sample selection bias, or may be from a different but related distribution. We present methods based on importance weighting that can estimate the loss with respect to a target distribution, even if we cannot access that distribution directly, in a variety of settings. These estimators, which differentially weight the contribution of data to the loss function, offer both theoretical guarantees and impressive empirical performance.