Confidence Regions for Predictions of Online Learning-Based Control. Issue 2 (2020)

• Record Type: Journal Article
• Authors: Capone, Alexandre; Lederer, Armin; Hirche, Sandra
• Is Part Of: IFAC-PapersOnLine. Volume 53:Issue 2(2020)
• Page Start: 1007
• Page End: 1012
• Publication Date: 2020
• Subjects: Gaussian processes -- system identification -- nonlinear systems -- stochastic systems -- Monte Carlo simulation -- error estimation; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2020.12.1278
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: Although machine learning techniques are increasingly employed in control tasks, few methods exist to predict the behavior of closed-loop learning-based systems. In this paper, we introduce a method for computing confidence regions of closed-loop system trajectories under an online learning-based control law. We employ a sampling-based approximation and exploit system properties to prove that the computed confidence regions are correct with high probability. In a numerical simulation, we show that the proposed approach accurately predicts correct confidence regions.