Title: Recent Advances and Progress in Evolutionary Algorithms and its Dynamics

Speaker: Prof. Ivan Zelinka

Location: Sala riunioni DINFO

This seminar is focused on recent progress in the mutual intersection of a few exciting fields of research whose core topic is evolutionary algorithms in general and will be explained relations between evolutionary dynamics, its visualization as a complex network and its control. It discusses research in evolutionary algorithms that can be considered a discrete dynamical complex system with inherent nonlinear dynamics. These dynamics is visualized and analyzed as a complex network and CML (coupled map lattices) systems. As already reported in many research papers and books, these dynamics can generate different kinds of behavior including chaotic one. Selected evolutionary algorithms discussed in this seminar will be differential evolution, genetic algorithm, particle swarm, artificial bee algorithm, and others. Methodology, converting evolutionary algorithms to the complex network will be introduced including demonstrations. The seminar then continues by explaining how evolutionary dynamics can be converted into so-called CML systems, which are used to model spatiotemporal behavior (including chaotic one) and also will be explained its relation to complex networks. At the end will be demonstrated how we can control EAs dynamics using a feedback loop control scheme. The seminar is designed as a topic overview, its structure, results, and content are based on our previous research, journal publications, keynotes, and tutorials and reflect the latest progress in that field.