The Information Bottleneck EM Algorithm

TL;DR

This paper introduces the IB-EM algorithm, which uses the Information Bottleneck principle to improve learning with hidden variables, overcoming local maxima issues of traditional EM by balancing information tradeoffs.

Contribution

The paper proposes a novel IB-EM algorithm that leverages information theory to enhance hidden variable learning, outperforming standard EM methods.

Findings

IB-EM finds better solutions than standard EM.

The approach balances informativeness and uninformative aspects of hidden variables.

It converges to high-scoring solutions by exploring information tradeoffs.

Abstract

Learning with hidden variables is a central challenge in probabilistic graphical models that has important implications for many real-life problems. The classical approach is using the Expectation Maximization (EM) algorithm. This algorithm, however, can get trapped in local maxima. In this paper we explore a new approach that is based on the Information Bottleneck principle. In this approach, we view the learning problem as a tradeoff between two information theoretic objectives. The first is to make the hidden variables uninformative about the identity of specific instances. The second is to make the hidden variables informative about the observed attributes. By exploring different tradeoffs between these two objectives, we can gradually converge on a high-scoring solution. As we show, the resulting, Information Bottleneck Expectation Maximization (IB-EM) algorithm, manages to find solutions that are superior to standard EM methods.