Training Set Debugging Using Trusted Items

TL;DR

This paper introduces a method to identify and correct bugs in training data using a small set of trusted items, improving model accuracy and trustworthiness.

Contribution

The paper presents a novel algorithm that formulates training set debugging as a bilevel optimization problem, effectively identifying label errors using trusted items.

Findings

Successfully detects training data bugs in toy and real datasets.

Proposes a relaxed continuous optimization approach for bug detection.

Enhances model reliability by correcting training set labels.

Abstract

Training set bugs are flaws in the data that adversely affect machine learning. The training set is usually too large for man- ual inspection, but one may have the resources to verify a few trusted items. The set of trusted items may not by itself be adequate for learning, so we propose an algorithm that uses these items to identify bugs in the training set and thus im- proves learning. Specifically, our approach seeks the smallest set of changes to the training set labels such that the model learned from this corrected training set predicts labels of the trusted items correctly. We flag the items whose labels are changed as potential bugs, whose labels can be checked for veracity by human experts. To find the bugs in this way is a challenging combinatorial bilevel optimization problem, but it can be relaxed into a continuous optimization problem. Ex- periments on toy and real data demonstrate that our approach can identify training set bugs effectively and suggest appro- priate changes to the labels. Our algorithm is a step toward trustworthy machine learning.