Participatory Search in Evolutionary Fuzzy Modeling.

Search is one of the most useful procedures employed in numerous situations such as optimization, machine learning, information processing and retrieval. This chapter introduces participatory search, a class of population-based search algorithms constructed upon the participatory learning paradigm. Participatory search relies on search mechanisms that progress forming pools of compatible individuals. The individual that is the most compatible with the best individual is always kept in the current population. Random immigrants are added to complete the population at each algorithm step. Different types of recombination are possible. The first is a convex combination, arithmetic-like recombination modulated by the compatibility between individuals. The second is a recombination mechanism based on selective transfer. Mutation is an instance of differential variation modulated by compatibility between selected and recombined individuals. Applications concerning development of fuzzy rule-based models from actual data illustrate the potential of the algorithms. The performance of the models produced by participatory search algorithms are compared with a state of the art genetic fuzzy system. Experimental results show that the participatory search algorithm with arithmetic-like recombination performs better than the remaining ones.