CUED Publications database

Nonlinear Set Membership Regression with Adaptive Hyper-Parameter Estimation for Online Learning and Control.

Calliess, JM and Roberts, S and Rasmussen, CE and Maciejowski, J (2018) Nonlinear Set Membership Regression with Adaptive Hyper-Parameter Estimation for Online Learning and Control. In: European Control Conference, 2018-6-12 to 2018-6-15, Limassol pp. 3167-3172..

Full text not available from this repository.

Abstract

Methods known as Lipschitz Interpolation or Nonlinear Set Membership regression have become established tools for nonparametric system-identification and data-based control. They utilise presupposed Lipschitz properties to compute inferences over unobserved function values. Unfortunately, it relies on the a priori knowledge of a Lipschitz constant of the underlying target function which serves as a hyperparameter. We propose a closed-form estimator of the Lipschitz constant that is robust to bounded observational noise in the data. The merger of Lipschitz Interpolation with the new hyperparameter estimator gives a new nonparametric machine learning method for which we derive sample complexity bounds and online learning convergence guarantees. Furthermore, we apply our learning method to model-reference adaptive control. We provide convergence guarantees on the closed-loop dynamics and compare the performance of our approach to recently proposed alternative learning-based controllers in a simulated flight manoeuvre control scenario.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Uncontrolled Keywords: iterative learning control, statistical learning adaptive control
Subjects: UNSPECIFIED
Divisions: Div F > Control
Div F > Computational and Biological Learning
Depositing User: Cron Job
Date Deposited: 10 Feb 2018 20:05
Last Modified: 12 Nov 2019 03:49
DOI: