Workshop

Update: the workshop program can be found here.

We are happy to invite you to the 8th Edition of the Workshop on Nonlinear System Identification Benchmarks held as an in-person event in Lugano, Switzerland, 24-26 April 2024. The workshop is preceded by a mini-course on nonlinear system identification, 22-23 April 2024.

This year again, the workshop is building upon its basic ingredients:

It stands as a forum facilitating the exchange of ideas and knowledge between the control, machine learning, and mechanical engineering communities.

A selection of renowned keynote speakers from the same 3 communities is invited to share with the audience their views on the future trends in the data-driven modeling for nonlinear dynamical systems. A tentative list of this year’s speakers can be found below:
- Melanie Zeilinger (ETH Zurich, Switzerland)
- Mario Sznaier (Northeastern University, USA)
- Roland Tóth (Eindhoven University of Technology, Netherlands & Sztaki, Hungary)
- Alberto Bemporad (IMT Lucca, Italy)

A 2-day course on deep learning for nonlinear system identification takes place before the workshop (22-23 April).

The workshop is open to both presenting and non-presenting participants. Contributions are expected to present the outcome of their favorite system identification approach on one or more of the benchmark datasets. The workshop is organized as an in-person event.

Important dates:

Early registration deadline: February 15th, 2024
Submission deadline: Friday, March 15, March 25, 2024 (please send a title + list of authors asap)
Workshop: 24-26 April 2024

Abstract submission:

Interested presenters are challenged to apply their favorite data-driven modeling tool to a carefully established list of benchmark systems and present their results as a regular presentation during the workshop. The list of benchmarks has recently been upgraded, and today 12 different benchmarks are proposed, addressing a wide range of state-of-the-art challenges in the field with various levels of difficulties.

Interested presenters can submit their contribution to m.schoukens@tue.nl as a 1-page abstract, following this template. Contributions are expected to present the outcome of their favorite system identification approach on one or more of the benchmark datasets that are listed on this webpage. Submissions that have no clear link to one of the featured datasets will most likely be rejected. We also welcome relevant contributions that have already been published elsewhere as the main goal of the workshop is to have interesting discussions and generate insights into the current practices and state-of-the-art of nonlinear system identification.

Information for presenters:

All accepted contributions will have a 30min timeslot to provide an oral presentation (20min presentation + 10min discussion). The workshop program, including presentation schedule will be announced soon.

Registration:

You can register through this link.

Early registration fees (before February 15th)