SUMMARY

Concepts and tools from statistical mechanics are nowadays used in a massive way in many different contexts, from fluid mechanics to biology and even in field far from physics. Moreover, recent advances in computational capabilities have opened up a large use of algorithms mainly based on information theory to build from data models, more or less extensively.

Yet, many of the present applications using statistical tools are far from the original framework of the statistical mechanics, and the possible impact of advanced computer science tools and most notably deep learning in building data-driven model is not yet clear. Such new approaches deserve an assessment. It would be important even to clarify which are the main issues to be addressed in the next years.

From the physical point of view, the workshop focuses on complex dynamical systems, with emphasis on turbulent ones, that is where spatio-temporal chaos is exhibited. Yet, recent developments in other non-equilibrium systems will be also considered. In particular, some issues in fields such collective motion, active matter and disordered systems are certainly relevant, since they have many common issues.

One goal of the workshop is to help the cross-fertilization of the different communities and to promote collaborations between them. A second purpose is to try to point out the most promising routes to solve interesting and relevant problems in the next future, possibly sharing the different and complementary knowledge.

In this respect, lectures will be given by people coming from different background, notably from fluid mechanics, non-equilibrium statistical physics, and machine-learnCing for physics.

All the participants will have the opportunity to present a poster.

We hope that the workshop will be particularly useful for the audience to advance their knowledge of this developping subject and to point out some new ideas about future studies to be carried out.