Adaptive Verifiable Coded Computing: Towards Fast, Secure and Private Distributed Machine Learning

TL;DR

This paper introduces AVCC, a flexible framework that separates handling of stragglers, privacy, and Byzantine faults in distributed machine learning, achieving significant speedups and accuracy improvements.

Contribution

AVCC decouples Byzantine detection from straggler tolerance, combining coded computing and verifiable computing for enhanced efficiency and adaptability.

Findings

Achieves up to 4.2x speedup over LCC

Improves accuracy by up to 5.1%

Speeds up distributed logistic regression by 7.6x

Abstract

Stragglers, Byzantine workers, and data privacy are the main bottlenecks in distributed cloud computing. Some prior works proposed coded computing strategies to jointly address all three challenges. They require either a large number of workers, a significant communication cost or a significant computational complexity to tolerate Byzantine workers. Much of the overhead in prior schemes comes from the fact that they tightly couple coding for all three problems into a single framework. In this paper, we propose Adaptive Verifiable Coded Computing (AVCC) framework that decouples the Byzantine node detection challenge from the straggler tolerance. AVCC leverages coded computing just for handling stragglers and privacy, and then uses an orthogonal approach that leverages verifiable computing to mitigate Byzantine workers. Furthermore, AVCC dynamically adapts its coding scheme to trade-off straggler tolerance with Byzantine protection. We evaluate AVCC on a compute-intensive distributed logistic regression application. Our experiments show that AVCC achieves up to $4.2\times$ speedup and up to $5.1\%$ accuracy improvement over the state-of-the-art Lagrange coded computing approach (LCC). AVCC also speeds up the conventional uncoded implementation of distributed logistic regression by up to $7.6\times$, and improves the test accuracy by up to $12.1\%$.