Modelling with Coupled Neuro-Fuzzy Map Lattices of Biological Processes
Technical University of Iasi and Institute for Computer Scie
Last modified: December 18, 2007
We investigate the nonlinear dynamics of fixed and adaptive coupled map lattices with neuro-fuzzy systems nodes. The lattices have various topologies, including rings with various vicinities, and linear, multilayered topologies. The use of fuzzy or neuro-fuzzy maps instead of deterministic (crisp) maps in coupled map lattices (CMLs) enhances the complexity of the models and makes them particularly useful for applications where imprecision and difficulty of grasping all the characteristics of the model are inherent, like in biology and sociology.
We use these models as tools to generate dynamic patterns close to those seen in various biological, medical, social, and economical populations. The results are commented in view of their relevance for deriving predictions on the future behaviours of such populations, when the initial state is known.