Inducing a hierarchy for multi-class classification problems

TL;DR

This paper proposes a method to induce hierarchical structures in flat multi-class classification problems, improving accuracy by clustering conditional distributions and using hierarchical classifiers, especially when no natural hierarchy exists.

Contribution

It introduces a novel approach to induce hierarchies in flat classification tasks, enhancing performance without requiring pre-existing label structures.

Findings

Induced hierarchies can improve classification accuracy.

Method effectively discovers latent hierarchies.

Approach outperforms flat classifiers in experiments.

Abstract

In applications where categorical labels follow a natural hierarchy, classification methods that exploit the label structure often outperform those that do not. Un-fortunately, the majority of classification datasets do not come pre-equipped with a hierarchical structure and classical flat classifiers must be employed. In this paper, we investigate a class of methods that induce a hierarchy that can similarly improve classification performance over flat classifiers. The class of methods follows the structure of first clustering the conditional distributions and subsequently using a hierarchical classifier with the induced hierarchy. We demonstrate the effectiveness of the class of methods both for discovering a latent hierarchy and for improving accuracy in principled simulation settings and three real data applications.