460-6017/02 – Bio-Inspired Computing (BIOIV)

Goals of the course: Bio-inspired computing

The content of the subject is following. Current state of the field of softcomputing, fuzzy logic, neural networks, evolutionary computing (EVT). Classification of evolutionary computational techniques, historical facts, current trends in EVT field. The central dogma of EVT by Darwin and Mendel. Basic concepts: individual, population, fitness, fitness function, representation of individuals. Fitness functions, design principles, test functions, computational complexity and theoretical limits of algorithms, P and NP problems. Permutation testing problems. Multipurpose optimization, Paret set, fitness function design for multipurpose optimization, examples. Selected stochastic algorithms: local search method, blind algorithm, climbing algorithm, simulated annealing. Selected stochastic algorithms with evolutionary elements: simulated annealing with elitism, taboo search. Particle swarm, Scatter Search, Ant Colony Optimization. Self-organizing Migration Algorithm, principle of operation and algorithm used: ATO, ATR, ATA and ATAA. SOMA and permutation test problems. Differential evolution.

Kenneth Price, Rainer M. Storn, Jouni A. Lampinen. Differential Evolution: A Practical Approach to Global Optimization, Springer, 2005. Christine Solnon. Ant Colony Optimization and Constraint Programming, Wiley-ISTE, 2010. Yang Xiao, Fei Hu. Bio-inspired Computing and Communication Networks, CRC, 2010.

Student vypracuje článek z vybraného pokročilého tématu. Tento článek odprezentuje v rámci kurzu. Ústní zkouška.