Towards Unbiased Exploration in Partial Label Learning

TL;DR

This paper addresses bias in neural partial label learning caused by softmax, proposing a new loss function that enables unbiased exploration and improves learning outcomes across various datasets.

Contribution

It introduces a novel loss function that corrects exploration bias in partial label learning, supported by theoretical analysis and extensive empirical evaluation.

Findings

The new loss reduces exploration bias in neural classifiers.

Improved performance on standard partial label benchmarks.

Effective in a novel rule learning benchmark.

Abstract

We consider learning a probabilistic classifier from partially-labelled supervision (inputs denoted with multiple possibilities) using standard neural architectures with a softmax as the final layer. We identify a bias phenomenon that can arise from the softmax layer in even simple architectures that prevents proper exploration of alternative options, making the dynamics of gradient descent overly sensitive to initialisation. We introduce a novel loss function that allows for unbiased exploration within the space of alternative outputs. We give a theoretical justification for our loss function, and provide an extensive evaluation of its impact on synthetic data, on standard partially labelled benchmarks and on a contributed novel benchmark related to an existing rule learning challenge.