One-Pass AUC Optimization

TL;DR

This paper introduces a novel one-pass AUC optimization algorithm that efficiently handles large, high-dimensional datasets by maintaining only essential statistics, enabling effective AUC maximization without multiple data passes.

Contribution

It proposes a regression-based, memory-efficient algorithm for one-pass AUC optimization that approximates covariance matrices for high-dimensional data.

Findings

Algorithm effectively optimizes AUC in a single pass.

Theoretical analysis confirms convergence and efficiency.

Empirical results demonstrate superior performance on large datasets.

Abstract

AUC is an important performance measure and many algorithms have been devoted to AUC optimization, mostly by minimizing a surrogate convex loss on a training data set. In this work, we focus on one-pass AUC optimization that requires only going through the training data once without storing the entire training dataset, where conventional online learning algorithms cannot be applied directly because AUC is measured by a sum of losses defined over pairs of instances from different classes. We develop a regression-based algorithm which only needs to maintain the first and second order statistics of training data in memory, resulting a storage requirement independent from the size of training data. To efficiently handle high dimensional data, we develop a randomized algorithm that approximates the covariance matrices by low rank matrices. We verify, both theoretically and empirically, the effectiveness of the proposed algorithm.