Analyzing the effect of prediction accuracy on the distributionally-robust competitive ratio

TL;DR

This paper investigates how prediction accuracy influences the distributionally-robust competitive ratio (DRCR) in algorithms with predictions, establishing its monotonic and concave relationship, and applies findings to the ski rental problem to determine accuracy thresholds.

Contribution

It proves that the optimal DRCR is a monotone and concave function of prediction accuracy and extends this analysis to multiple predictions, providing insights into performance guarantees.

Findings

Optimal DRCR decreases linearly with prediction accuracy.

Optimal DRCR is a monotone and concave function of accuracy.

Conditions for achieving target DRCR in ski rental problem.

Abstract

The field of algorithms with predictions aims to improve algorithm performance by integrating machine learning predictions into algorithm design. A central question in this area is how predictions can improve performance, and a key aspect of this analysis is the role of prediction accuracy. In this context, prediction accuracy is defined as a guaranteed probability that an instance drawn from the distribution belongs to the predicted set. As a performance measure that incorporates prediction accuracy, we focus on the distributionally-robust competitive ratio (DRCR), introduced by Sun et al.~(ICML 2024). The DRCR is defined as the expected ratio between the algorithm's cost and the optimal cost, where the expectation is taken over the worst-case instance distribution that satisfies the given prediction and accuracy requirement. A known structural property is that, for any fixed algorithm, the DRCR decreases linearly as prediction accuracy increases. Building on this result, we establish that the optimal DRCR value (i.e., the infimum over all algorithms) is a monotone and concave function of prediction accuracy. We further generalize the DRCR framework to a multiple-prediction setting and show that monotonicity and concavity are preserved in this setting. Finally, we apply our results to the ski rental problem, a benchmark problem in online optimization, to identify the conditions on prediction accuracies required for the optimal DRCR to attain a target value. Moreover, we provide a method for computing the critical accuracy, defined as the minimum accuracy required for the optimal DRCR to strictly improve upon the performance attainable without any accuracy guarantee.