Koopman Spectrum Nonlinear Regulators and Efficient Online Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ohnishi, Motoya, Ishikawa, Isao, Lowrey, Kendall, Ikeda, Masahiro, Kakade, Sham, Kawahara, Yoshinobu
Format: Preprint
Published: 2021
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913412067360768
author Ohnishi, Motoya
Ishikawa, Isao
Lowrey, Kendall
Ikeda, Masahiro
Kakade, Sham
Kawahara, Yoshinobu
author_facet Ohnishi, Motoya
Ishikawa, Isao
Lowrey, Kendall
Ikeda, Masahiro
Kakade, Sham
Kawahara, Yoshinobu
contents Most modern reinforcement learning algorithms optimize a cumulative single-step cost along a trajectory. The optimized motions are often 'unnatural', representing, for example, behaviors with sudden accelerations that waste energy and lack predictability. In this work, we present a novel paradigm of controlling nonlinear systems via the minimization of the Koopman spectrum cost: a cost over the Koopman operator of the controlled dynamics. This induces a broader class of dynamical behaviors that evolve over stable manifolds such as nonlinear oscillators, closed loops, and smooth movements. We demonstrate that some dynamics characterizations that are not possible with a cumulative cost are feasible in this paradigm, which generalizes the classical eigenstructure and pole assignments to nonlinear decision making. Moreover, we present a sample efficient online learning algorithm for our problem that enjoys a sub-linear regret bound under some structural assumptions.
format Preprint
id arxiv_https___arxiv_org_abs_2106_15775
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Koopman Spectrum Nonlinear Regulators and Efficient Online Learning
Ohnishi, Motoya
Ishikawa, Isao
Lowrey, Kendall
Ikeda, Masahiro
Kakade, Sham
Kawahara, Yoshinobu
Machine Learning
Robotics
Systems and Control
Most modern reinforcement learning algorithms optimize a cumulative single-step cost along a trajectory. The optimized motions are often 'unnatural', representing, for example, behaviors with sudden accelerations that waste energy and lack predictability. In this work, we present a novel paradigm of controlling nonlinear systems via the minimization of the Koopman spectrum cost: a cost over the Koopman operator of the controlled dynamics. This induces a broader class of dynamical behaviors that evolve over stable manifolds such as nonlinear oscillators, closed loops, and smooth movements. We demonstrate that some dynamics characterizations that are not possible with a cumulative cost are feasible in this paradigm, which generalizes the classical eigenstructure and pole assignments to nonlinear decision making. Moreover, we present a sample efficient online learning algorithm for our problem that enjoys a sub-linear regret bound under some structural assumptions.
title Koopman Spectrum Nonlinear Regulators and Efficient Online Learning
topic Machine Learning
Robotics
Systems and Control
url https://arxiv.org/abs/2106.15775