Efficient Numerical Analysis in Complexity Science

International Workshop
19 - 23 May 2025

The study of complex systems is almost impossible without numerical
simulations and analysis. Consequently, progress in this field has been and
will continue to be closely linked to progress in the field of computational
physics. Despite great breakthroughs also in analytical understanding, most
insights started from numerical observations. Famous examples of physical
phenomena, methods, and tools include the Fermi-Pasta-Ulam problem, Monte
Carlo simulations, percolation, diffusion limited aggregation, deterministic
chaos, fractals, stochastic resonance, self-organized-criticality, rare
event simulation, modeling infectious diseases, pattern formation, and
measuring mutual information in time series data, just to mention a few. In
all these examples, large scale numerics is needed in one or another
way. Numerical analysis also provides the link between observed data and models.
While the available compute power has witnessed an increase by about 9
orders of magnitude in the past 40 years, the ambitions of scientists in terms
of statistical accuracy and system size have increased even faster. Therefore
not only in the early times but still today, efficient algorithms are
essential.

This workshop will review past achievements, novel approaches, and future
directions in computational aspects of complexity science and their
applications to specific systems. It will be held
also in honor of Peter Grassberger, who has tremendeously
driven this field in the past 50 years, and who will celebrate his 85th
anniversary in 2025.

• Percolation theory
• Sampling techniques of rare events
• Data analysis with fast and stable algorithms
• Statistical inference of power laws and their physical origins
• Sub-diffusion and ultra-slow diffusion
• Machine learning from the dynamical systems' point of view
• Networks and dynamics on networks