English

Learning Linear Models Using Distributed Iterative Hessian Sketching

Optimization and Control 2021-12-09 v1 Machine Learning Numerical Analysis Systems and Control Systems and Control Numerical Analysis

Abstract

This work considers the problem of learning the Markov parameters of a linear system from observed data. Recent non-asymptotic system identification results have characterized the sample complexity of this problem in the single and multi-rollout setting. In both instances, the number of samples required in order to obtain acceptable estimates can produce optimization problems with an intractably large number of decision variables for a second-order algorithm. We show that a randomized and distributed Newton algorithm based on Hessian-sketching can produce ϵ\epsilon-optimal solutions and converges geometrically. Moreover, the algorithm is trivially parallelizable. Our results hold for a variety of sketching matrices and we illustrate the theory with numerical examples.

Keywords

Cite

@article{arxiv.2112.04101,
  title  = {Learning Linear Models Using Distributed Iterative Hessian Sketching},
  author = {Han Wang and James Anderson},
  journal= {arXiv preprint arXiv:2112.04101},
  year   = {2021}
}
R2 v1 2026-06-24T08:08:32.787Z