Unbiased Loss Functions for Multilabel Classification with Missing Labels

TL;DR

This paper develops unbiased loss functions for multilabel classification with missing labels, addressing the bias introduced by missing data and proposing solutions that balance bias and variance in training.

Contribution

It derives the unique unbiased estimators for multilabel reductions, including non-decomposable ones, and proposes convex upper-bounds to mitigate variance issues.

Findings

Unbiased estimators significantly change the bias-variance trade-off.

Switching to unbiased estimators may require stronger regularization.

Unbiased estimators can improve training in missing-label multilabel tasks.

Abstract

This paper considers binary and multilabel classification problems in a setting where labels are missing independently and with a known rate. Missing labels are a ubiquitous phenomenon in extreme multi-label classification (XMC) tasks, such as matching Wikipedia articles to a small subset out of the hundreds of thousands of possible tags, where no human annotator can possibly check the validity of all the negative samples. For this reason, propensity-scored precision -- an unbiased estimate for precision-at-k under a known noise model -- has become one of the standard metrics in XMC. Few methods take this problem into account already during the training phase, and all are limited to loss functions that can be decomposed into a sum of contributions from each individual label. A typical approach to training is to reduce the multilabel problem into a series of binary or multiclass problems, and it has been shown that if the surrogate task should be consistent for optimizing recall, the resulting loss function is not decomposable over labels. Therefore, this paper derives the unique unbiased estimators for the different multilabel reductions, including the non-decomposable ones. These estimators suffer from increased variance and may lead to ill-posed optimization problems, which we address by switching to convex upper-bounds. The theoretical considerations are further supplemented by an experimental study showing that the switch to unbiased estimators significantly alters the bias-variance trade-off and may thus require stronger regularization, which in some cases can negate the benefits of unbiased estimation.