Adversarially Robust Multitask Adaptive Control

TL;DR

This paper introduces a robust multitask adaptive control method that combines clustering, system identification, and resilient aggregation to improve control policy learning under adversarial conditions and model uncertainty.

Contribution

It proposes a novel clustered multitask approach for adversarially robust control that effectively mitigates corrupted updates and provides theoretical regret bounds.

Findings

Regret decreases inversely with honest systems per cluster

Method remains effective under bounded adversarial systems

Non-asymptotic bounds demonstrate robustness and efficiency

Abstract

We study adversarially robust multitask adaptive linear quadratic control; a setting where multiple systems collaboratively learn control policies under model uncertainty and adversarial corruption. We propose a clustered multitask approach that integrates clustering and system identification with resilient aggregation to mitigate corrupted model updates. Our analysis characterizes how clustering accuracy, intra-cluster heterogeneity, and adversarial behavior affect the expected regret of certainty-equivalent (CE) control across LQR tasks. We establish non-asymptotic bounds demonstrating that the regret decreases inversely with the number of honest systems per cluster and that this reduction is preserved under a bounded fraction of adversarial systems within each cluster.