Tree-structure based hybrid computational intelligence. Theoretical foundations and applications (Q1046720)

The book puts forward a hierarchical hybrid computational intelligence framework in which models of a hierarchical structure and appropriate types of function operators are created and optimized by means of computational intelligence techniques. Optimization alternates tree structure evolution with parameter setting until either an optimal solution or a maximum time limit is reached. In this respect, Chapters 2, 3 and 4 outline such variants of hybrid computational intelligence models for static environments, e.g. neural trees, radial basis function networks, B-spline networks, wavelet neural networks, fuzzy systems. The focus is then changed to dynamic systems in Chapter 5, when additive tree models are presented. Finally, Chapter 6 integrates all the concepts and mechanisms into the proposed hierarchical hybrid computational intelligence framework. A large and varied set of real-world applications is outlined in connection to each addressed component, e.g. stock market prediction, intrusion detection, exchange rate forecasting, face recognition, time series prediction, system identification, breast cancer detection and protein fold recognition. The book provides some concise introduction to the founding paradigms of computational intelligence, but mostly offers ideas and prospects for current practical research interests and challenges, thus making itself more appropriate for experienced researchers in the field. The book is both informative and stimulating and thus becomes an interesting and valuable source.