The Nonstochastic Control Problem

Elad Hazan, Sham M. Kakade, Karan Singh

Research output: Contribution to journalConference articlepeer-review

Abstract

We consider the problem of controlling an unknown linear dynamical system in the presence of (nonstochastic) adversarial perturbations and adversarial convex loss functions. In contrast to classical control, the a priori determination of an optimal controller here is hindered by the latter’s dependence on the yet unknown perturbations and costs. Instead, we measure regret against an optimal linear policy in hindsight, and give the first efficient algorithm that guarantees a sublinear regret bound, scaling as O(T2/3), in this setting.

Original languageEnglish (US)
Pages (from-to)408-421
Number of pages14
JournalProceedings of Machine Learning Research
Volume117
StatePublished - 2020
Event31st International Conference on Algorithmic Learning Theory, ALT 2020 - San Diego, United States
Duration: Feb 8 2020Feb 11 2020

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software
  • Control and Systems Engineering
  • Statistics and Probability

Keywords

  • Control Theory
  • Online Learning
  • Robust Control
  • System Identification

Fingerprint

Dive into the research topics of 'The Nonstochastic Control Problem'. Together they form a unique fingerprint.

Cite this