A submodular-supermodular procedure with applications to discriminative structure learning

TL;DR

This paper introduces a variational algorithm for minimizing the difference between two submodular functions, with applications in discriminative structure learning and feature selection, demonstrating improved classifier performance.

Contribution

The paper proposes a novel submodular-supermodular procedure extending the concave-convex method, tailored for machine learning tasks involving submodular metrics.

Findings

The algorithm effectively minimizes submodular differences in polynomial time.

Discriminative graphical models using this method outperform generative models.

Synthetic data experiments validate the approach's effectiveness.

Abstract

In this paper, we present an algorithm for minimizing the difference between two submodular functions using a variational framework which is based on (an extension of) the concave-convex procedure [17]. Because several commonly used metrics in machine learning, like mutual information and conditional mutual information, are submodular, the problem of minimizing the difference of two submodular problems arises naturally in many machine learning applications. Two such applications are learning discriminatively structured graphical models and feature selection under computational complexity constraints. A commonly used metric for measuring discriminative capacity is the EAR measure which is the difference between two conditional mutual information terms. Feature selection taking complexity considerations into account also fall into this framework because both the information that a set of features provide and the cost of computing and using the features can be modeled as submodular functions. This problem is NP-hard, and we give a polynomial time heuristic for it. We also present results on synthetic data to show that classifiers based on discriminative graphical models using this algorithm can significantly outperform classifiers based on generative graphical models.