Towards Interpretable Deep Extreme Multi-label Learning

TL;DR

This paper addresses the interpretability challenge in deep extreme multi-label learning by proposing a two-step approach that combines deep autoencoders with classifiers, enabling understanding of label hierarchies and dependencies.

Contribution

It introduces a novel two-step XML method integrating deep non-negative autoencoders with classifiers for improved interpretability and scalability in multi-label learning.

Findings

Effective handling of large label sets

Provides interpretable label hierarchies

Enhances understanding of model decision processes

Abstract

Many Machine Learning algorithms, such as deep neural networks, have long been criticized for being "black-boxes"-a kind of models unable to provide how it arrive at a decision without further efforts to interpret. This problem has raised concerns on model applications' trust, safety, nondiscrimination, and other ethical issues. In this paper, we discuss the machine learning interpretability of a real-world application, eXtreme Multi-label Learning (XML), which involves learning models from annotated data with many pre-defined labels. We propose a two-step XML approach that combines deep non-negative autoencoder with other multi-label classifiers to tackle different data applications with a large number of labels. Our experimental result shows that the proposed approach is able to cope with many-label problems as well as to provide interpretable label hierarchies and dependencies that helps us understand how the model recognizes the existences of objects in an image.