Neuro-fuzzy structures and their comparative analyzes
Robert K. Nowicki, Agata Pokropińska · 2005
In the paper various neuro-fuzzy architectures will be tested and compared. These architectures are based on the weighted version of triangular norms in rule aggregation and Cartesian product parts. The main goal of the paper is to investigate if the new architectures lead to improvement of results of classification and approximation. The structures will be tested on the typical benchmarks (Wine Recognition, Wisconsin Breast Cancer Data, Nonlinear Dynamic Plant and Modeling of a Static Nonlinear Function).