No-Regret Caching with Noisy Request Estimates

TL;DR

This paper introduces the NFPL algorithm, a novel online caching strategy that operates effectively with noisy request data, achieving low regret in high load or memory-constrained environments.

Contribution

It develops the NFPL algorithm, extending Follow-the-Perturbed-Leader to handle noisy request estimates with proven regret guarantees.

Findings

NFPL outperforms classic policies on synthetic traces.

NFPL performs comparably to optimal policies on real request data.

The approach is effective under specific conditions on request estimators.

Abstract

Online learning algorithms have been successfully used to design caching policies with regret guarantees. Existing algorithms assume that the cache knows the exact request sequence, but this may not be feasible in high load and/or memory-constrained scenarios, where the cache may have access only to sampled requests or to approximate requests' counters. In this paper, we propose the Noisy-Follow-the-Perturbed-Leader (NFPL) algorithm, a variant of the classic Follow-the-Perturbed-Leader (FPL) when request estimates are noisy, and we show that the proposed solution has sublinear regret under specific conditions on the requests estimator. The experimental evaluation compares the proposed solution against classic caching policies and validates the proposed approach under both synthetic and real request traces.