Combating the Instability of Mutual Information-based Losses via Regularization

arXiv2011.07932MaRDI QIDQ6353838

Author name not available (Why is that?)

Publication date: 16 November 2020

Abstract: Notable progress has been made in numerous fields of machine learning based on neural network-driven mutual information (MI) bounds. However, utilizing the conventional MI-based losses is often challenging due to their practical and mathematical limitations. In this work, we first identify the symptoms behind their instability: (1) the neural network not converging even after the loss seemed to converge, and (2) saturating neural network outputs causing the loss to diverge. We mitigate both issues by adding a novel regularization term to the existing losses. We theoretically and experimentally demonstrate that added regularization stabilizes training. Finally, we present a novel benchmark that evaluates MI-based losses on both the MI estimation power and its capability on the downstream tasks, closely following the pre-existing supervised and contrastive learning settings. We evaluate six different MI-based losses and their regularized counterparts on multiple benchmarks to show that our approach is simple yet effective.

Has companion code repository: https://github.com/siyeong-lee/deconstructing-mine

This page was built for publication: Combating the Instability of Mutual Information-based Losses via Regularization

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6353838)