The project will develop methods and prototype software tools for modeling and managing pattern resilience in complex systems. Pattern resilience is understood as the capacity of the system to maintain or to recover some desired pattern dynamics in a changing environment. The patterns statistical regularities which are generated by the interconnected components of the system. They evolve with time which justifies the term 'pattern dynamics'.
We shall build on statistical physics and Pattern Oriented Modeling techniques to develop a set of robust techniques to express such pattern dynamics in state space of relatively small dimensionality. Viability theory will provide a framework for an inclusive, rigorous and practical definition of resilience, which give possibilities to find sets of actions to maintain or restore the satisfactory pattern dynamics. We shall test the methods and the software tools on a set of concrete applications, in collaboration with field specialists.
We shall build on the tests to refine and improve the methods and tools. The applications include: bacteria dynamics, land-use in semi-arid savannas, learning of sequences in basal ganglia, language variety, biotech firm networks. We expect the use of these new methods to bring new knowledge about the resilience of these systems.
Field of science
- /natural sciences/computer and information sciences/software
Call for proposal
See other projects for this call