Skip to Main content Skip to Navigation
New interface
Conference papers

Online Learning and Control of Dynamical Systems from Sensory Input

Abstract : Identifying an effective model of a dynamical system from sensory data and using it for future state prediction and control is challenging. Recent data-driven algorithms based on Koopman theory are a promising approach to this problem, but they typically never update the model once it has been identified from a relatively small set of observations, thus making long-term prediction and control difficult for realistic systems, in robotics or fluid mechanics for example. This paper introduces a novel method for learning an embedding of the state space with linear dynamics from sensory data. Unlike previous approaches, the dynamics model can be updated online and thus easily applied to systems with non-linear dynamics in the original configuration space. The proposed approach is evaluated empirically on several classical dynamical systems and sensory modalities, with good performance on long-term prediction and control.
Complete list of metadata
Contributor : Oumayma Bounou Connect in order to contact the contributor
Submitted on : Sunday, November 7, 2021 - 9:54:57 PM
Last modification on : Saturday, November 19, 2022 - 3:59:02 AM


Files produced by the author(s)


  • HAL Id : hal-03405911, version 2


Oumayma Bounou, Jean Ponce, Justin Carpentier. Online Learning and Control of Dynamical Systems from Sensory Input. NeurIPS 2021 - Thirty-fifth Conference on Neural Information Processing Systems Year, Dec 2021, Sydney / Virtual, Australia. ⟨hal-03405911v2⟩



Record views


Files downloads