A unified framework for hard and soft clustering with regularized optimal transport

TL;DR

This paper introduces a unified clustering framework using regularized optimal transport that encompasses hard and soft clustering, with a parameter controlling the nature of the clustering, and provides convergence analysis and empirical benefits.

Contribution

The paper presents a novel optimal transport-based clustering method that unifies hard and soft clustering, including a generalized EM algorithm with convergence guarantees.

Findings

Using b1>1 improves inference performance.

Approaching b1a0bca0bca0bca0bc enhances inference accuracy.

Approaching b1a0bca0bca0bca0bc improves classification.

Abstract

In this paper, we formulate the problem of inferring a Finite Mixture Model from discrete data as an optimal transport problem with entropic regularization of parameter $\lambda\geq 0$. Our method unifies hard and soft clustering, the Expectation-Maximization (EM) algorithm being exactly recovered for $\lambda=1$. The family of clustering algorithm we propose rely on the resolution of nonconvex problems using alternating minimization. We study the convergence property of our generalized $\lambda-$EM algorithms and show that each step in the minimization process has a closed form solution when inferring finite mixture models of exponential families. Experiments highlight the benefits of taking a parameter $\lambda>1$ to improve the inference performance and $\lambda\to 0$ for classification.