arXiv subjects

Peter Coppens

Publications and source records attributed to Peter Coppens.

6 recordsLinked to original sources

arXiv2023

Ordered Risk Minimization: Learning More from Less Data

We consider the worst-case expectation of a permutation invariant ambiguity set of discrete distributions as a proxy-cost for data-driven expected risk minimization. For this framework, we coin the term ordered risk minimization to highlight how results from order statistics inspired the proxy-cost. Specifically, we show how such costs serve as point-wise high-confidence upper bounds of the expected risk. The confidence level can be determined tightly for any sample size. Conversely we also illustrate how to calibrate the size of the ambiguity set such that the high-confidence upper bound has some user specified confidence. This calibration procedure notably supports $\phi$-divergence based ambiguity sets. Numerical experiments then illustrate how the resulting scheme both generalizes better and is less sensitive to tuning parameters compared to the empirical risk minimization approach.

math.OC↗

arXiv2022

Policy iteration using Q-functions: Linear dynamics with multiplicative noise

This paper presents a novel model-free and fully data-driven policy iteration scheme for quadratic regulation of linear dynamics with state- and input-multiplicative noise. The implementation is similar to the least-squares temporal difference scheme for Markov decision processes, estimating Q-functions by solving a least-squares problem with instrumental variables. The scheme is compared with a model-based system identification scheme and natural policy gradient through numerical experiments.

math.OC↗

arXiv2022

Provably stable learning control of linear dynamics with multiplicative noise

Control of linear dynamics with multiplicative noise naturally introduces robustness against dynamical uncertainty. Moreover, many physical systems are subject to multiplicative disturbances. In this work we show how these dynamics can be identified from state trajectories. The least-squares scheme enables exploitation of prior information and comes with practical data-driven confidence bounds and sample complexity guarantees. We complement this scheme with an associated control synthesis procedure for LQR which robustifies against distributional uncertainty, guaranteeing stability with high probability and converging to the true optimum at a rate inversely proportional with the sample count. Throughout we exploit the underlying multi-linear problem structure through tensor algebra and completely positive operators. The scheme is validated through numerical experiments.

math.OC↗

arXiv2021

Data-driven distributionally robust MPC for constrained stochastic systems

In this paper we introduce a novel approach to distributionally robust optimal control that supports online learning of the ambiguity set, while guaranteeing recursive feasibility. We introduce conic representable risk, which is useful to derive tractable reformulations of distributionally robust optimization problems. Specifically, to illustrate the techniques introduced, we utilize risk measures constructed based on data-driven ambiguity sets, constraining the second moment of the random disturbance. In the optimal control setting, such moment-based risk measures lead to tractable optimal controllers when combined with affine disturbance feedback. Assumptions on the constraints are given that guarantee recursive feasibility. The resulting control scheme acts as a robust controller when little data is available and converges to the certainty equivalent controller when a large sample count implies high confidence in the estimated second moment. This is illustrated in a numerical experiment.

eess.SY↗

arXiv2020

Sample Complexity of Data-Driven Stochastic LQR with Multiplicative Uncertainty

This paper studies the sample complexity of the stochastic Linear Quadratic Regulator when applied to systems with multiplicative noise. We assume that the covariance of the noise is unknown and estimate it using the sample covariance, which results in suboptimal behaviour. The main contribution of this paper is then to bound the suboptimality of the methodology and prove that it decreases with 1/N, where N denotes the amount of samples. The methodology easily generalizes to the case where the mean is unknown and to the distributionally robust case studied in a previous work of the authors. The analysis is mostly based on results from matrix function perturbation analysis.

eess.SY↗

arXiv2019

Data-driven distributionally robust LQR with multiplicative noise

We present a data-driven method for solving the linear quadratic regulator problem for systems with multiplicative disturbances, the distribution of which is only known through sample estimates. We adopt a distributionally robust approach to cast the controller synthesis problem as semidefinite programs. Using results from high dimensional statistics, the proposed methodology ensures that their solution provides mean-square stabilizing controllers with high probability even for low sample sizes. As sample size increases the closed-loop cost approaches that of the optimal controller produced when the distribution is known. We demonstrate the practical applicability and performance of the method through a numerical experiment.

eess.SY↗