Robustifying Learning-Augmented Caching Efficiently without Compromising 1-Consistency

TL;DR

This paper introduces Guard, a lightweight framework that improves robustness of learning-augmented caching algorithms to a factor of approximately 2, without losing their optimal 1-consistency, and with minimal computational overhead.

Contribution

The paper presents Guard, a novel robustification framework that enhances robustness to a factor of about 2H_{k-1}+2 while maintaining 1-consistency and low overhead.

Findings

Guard achieves the best trade-off between robustness and consistency.

Extensive experiments validate Guard's effectiveness in real-world scenarios.

Guard introduces only O(1) additional overhead per request.

Abstract

The online caching problem aims to minimize cache misses when serving a sequence of requests under a limited cache size. While naive learning-augmented caching algorithms achieve ideal $1$-consistency, they lack robustness guarantees. Existing robustification methods either sacrifice $1$-consistency or introduce excessive computational overhead. In this paper, we introduce Guard, a lightweight robustification framework that enhances the robustness of a broad class of learning-augmented caching algorithms to $2H_{k-1} + 2$, while preserving their $1$-consistency. Guard achieves the current best-known trade-off between consistency and robustness, with only O(1) additional per-request overhead, thereby maintaining the original time complexity of the base algorithm. Extensive experiments across multiple real-world datasets and prediction models validate the effectiveness of Guard in practice.