Header logo is

Actively Learning Gaussian Process Dynamics


Technical Report


Despite the availability of ever more data enabled through modern sensor and computer technology, it still remains an open problem to learn dynamical systems in a sample-efficient way. We propose active learning strategies that leverage information-theoretical properties arising naturally during Gaussian process regression, while respecting constraints on the sampling process imposed by the system dynamics. Sample points are selected in regions with high uncertainty, leading to exploratory behavior and data-efficient training of the model. All results are verified in an extensive numerical benchmark.

Author(s): Mona Buisson-Fenet and Friedrich Solowjow and Sebastian Trimpe
Year: 2019

Department(s): Intelligent Control Systems
Bibtex Type: Technical Report (techreport)
Paper Type: Work in Progress

State: Submitted

Links: ArXiv


@techreport{Actively Learning Gaussian Process Dynamics,
  title = {Actively Learning Gaussian Process Dynamics},
  author = {Buisson-Fenet, Mona and Solowjow, Friedrich and Trimpe, Sebastian},
  year = {2019}