Multiple Importance Sampling for Stochastic Gradient Estimation

TL;DR

This paper presents a novel importance sampling framework for stochastic gradient estimation that adaptively combines multiple distributions to improve training efficiency and convergence across diverse tasks.

Contribution

It introduces a self-adaptive, multi-distribution importance sampling method for gradient estimation, enhancing convergence speed and accuracy in training deep models.

Findings

Improved gradient estimates lead to faster training convergence.

Effective across classification, regression, image, and point cloud tasks.

Outperforms traditional sampling methods in empirical evaluations.

Abstract

We introduce a theoretical and practical framework for efficient importance sampling of mini-batch samples for gradient estimation from single and multiple probability distributions. To handle noisy gradients, our framework dynamically evolves the importance distribution during training by utilizing a self-adaptive metric. Our framework combines multiple, diverse sampling distributions, each tailored to specific parameter gradients. This approach facilitates the importance sampling of vector-valued gradient estimation. Rather than naively combining multiple distributions, our framework involves optimally weighting data contribution across multiple distributions. This adapted combination of multiple importance yields superior gradient estimates, leading to faster training convergence. We demonstrate the effectiveness of our approach through empirical evaluations across a range of optimization tasks like classification and regression on both image and point cloud datasets.