Finding High-Value Training Data Subset through Differentiable Convex Programming

TL;DR

This paper introduces a scalable, learnable framework for selecting high-value training data subsets for deep neural networks, leveraging differentiable convex programming to improve subset quality and identify mislabeled data.

Contribution

It proposes a novel end-to-end learnable method for subset selection that accounts for data point interactions, outperforming existing valuation techniques.

Findings

Achieves up to 20% higher subset value than state-of-the-art methods.

Effectively identifies mislabeled training data.

Runs with comparable efficiency to existing valuation functions.

Abstract

Finding valuable training data points for deep neural networks has been a core research challenge with many applications. In recent years, various techniques for calculating the "value" of individual training datapoints have been proposed for explaining trained models. However, the value of a training datapoint also depends on other selected training datapoints - a notion that is not explicitly captured by existing methods. In this paper, we study the problem of selecting high-value subsets of training data. The key idea is to design a learnable framework for online subset selection, which can be learned using mini-batches of training data, thus making our method scalable. This results in a parameterized convex subset selection problem that is amenable to a differentiable convex programming paradigm, thus allowing us to learn the parameters of the selection model in end-to-end training. Using this framework, we design an online alternating minimization-based algorithm for jointly learning the parameters of the selection model and ML model. Extensive evaluation on a synthetic dataset, and three standard datasets, show that our algorithm finds consistently higher value subsets of training data, compared to the recent state-of-the-art methods, sometimes ~20% higher value than existing methods. The subsets are also useful in finding mislabelled training data. Our algorithm takes running time comparable to the existing valuation functions.