Learning Diverse and Discriminative Representations via the Principle of Maximal Coding Rate Reduction

arXiv2006.08558MaRDI QIDQ6342974

Author name not available (Why is that?)

Publication date: 15 June 2020

Abstract: To learn intrinsic low-dimensional structures from high-dimensional data that most discriminate between classes, we propose the principle of Maximal Coding Rate Reduction ($ext{MCR}^2$), an information-theoretic measure that maximizes the coding rate difference between the whole dataset and the sum of each individual class. We clarify its relationships with most existing frameworks such as cross-entropy, information bottleneck, information gain, contractive and contrastive learning, and provide theoretical guarantees for learning diverse and discriminative features. The coding rate can be accurately computed from finite samples of degenerate subspace-like distributions and can learn intrinsic representations in supervised, self-supervised, and unsupervised settings in a unified manner. Empirically, the representations learned using this principle alone are significantly more robust to label corruptions in classification than those using cross-entropy, and can lead to state-of-the-art results in clustering mixed data from self-learned invariant features.

Has companion code repository: https://github.com/Ma-Lab-Berkeley/MCR2

This page was built for publication: Learning Diverse and Discriminative Representations via the Principle of Maximal Coding Rate Reduction

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6342974)