Sanchez D., Melin P. Hierarchical Modular Granular Neural Networks with Fuzzy Aggregation

Springer, 2017. — 105 p.In this book, a new model of modular neural network based on a granular approach, the combination of their responses, and the optimization by hierarchical genetic algorithms are introduced. The new model of modular neural networks is applied to human recognition, and for this four databases of biometric measures are used; face, iris, ear, and voice. The different responses are combined using type-1 and interval type-2 fuzzy logic. Finally, two hierarchical genetic algorithms are used to perform the optimization of the granular modular neural networks parameters and fuzzy inference system parameters. The experimental results obtained using the proposed method show that when the optimization is used, the results can be better than without optimization.
This book is intended to be a reference for scientists and engineers interested in applying soft computing techniques, such as neural networks, fuzzy logic, and genetic algorithms; all of them apply to human recognition, but also in general to pattern recognition and hybrid intelligent systems and similar ones. We consider that this book can also be used to find novel ideas for new lines of research, or to continue the lines of research proposed by authors of the book.Background and Theory
Proposed Method
Application to Human Recognition
Experimental ResultsBiometric Measures
Fuzzy Integration