Learning Chaotic Dynamics:
from Toy Models to Real World Applications

International Focus Workshop
1 - 3 March 2027


Machine learning nowadays is ubiquitous. In the natural sciences, it has become an important tool for analysis, classification, and prediction. Out of the huge breadth of topics, this workshop focuses on the aspect of learning deterministic but chaotic dynamics from data or to extract deterministic features from seemingly stochastic processes. The meeting will span multiple machine learning approaches, from recent foundation models to reservoir computing (both digital and physical).

Topics include

  • comparison of different approaches/architectures
  • hyper-parameter optimization
  • short term predictions and long term modeling and stability
  • merging output of multiple models
  • robustness to noise
  • applications to time series data from nature and experiments

Invited speakers

Manuel Brenner (DE)
Claudio Gallicchio (IT)
Lyudmila Grigoryeva (CH)
Jason Z. Kim (US)
Ying-Cheng Lai (US)
Luca Magri* (UK)
Zahra Monfared (DE)
Ulrich Parlitz (DE)
Chris Rackauckas* (US)
Pantelis Rafail Vlachas* (CH)

* to be confirmed

 

Scientific Coordinators

Daniel Durstewitz
Central Institute of Mental Health, Mannheim, Germany

Holger Kantz
MPIPKS, Dresden, Germany

Organisation

Katrin Lantsch
MPIPKS Dresden, Germany

Download

image.alternative

Application

The application deadline for LCD27 is 6th December 2026. The number of attendees is limited.

The registration fee for the international focus workshop is 200 Euro and must be paid by all participants.

Costs for accommodation and meals will be covered by the Max Planck Institute.

Limited funding is available to partially cover travel expenses.

Scientific Program

Program structure:

Sunday, 28th February - arrival (recommended)

Monday, 1st March - 10 am program start

Wednesday, 3rd March - 1 pm program closure and departure

The detailed program will be available by mid February 2027, approximately.

Accommodation Information

Please beware of accommodation arrangements and other services offered by third parties. We urge you not to give your credit card details or other personal information. The organization of the event is solely handled by the Visitors Program of MPI-PKS Dresden and its staff, and any communication on the topic of accommodation will always come from @pks.mpg.de addresses or from the scientific coordinators. If in doubt, please contact us via the event email address or via visitors[at]pks.mpg(dot)de. Thank you!