Data-driven initialization and structure learning in fuzzy neural networks

Initialization and structure learning in fuzzy neural networks for data-driven rule-based modeling are discussed. Gradient-based optimization is used to fit the model to data and a number of techniques are developed to enhance transparency of the generated rule base: data-driven initialization, similarity analysis for redundancy reduction, and evaluation of the rules contributions. The initialization uses flexible hyper-boxes to avoid redundant and irrelevant coverage of the input space. Similarity analysis detects redundant terms while the contribution evaluation detects irrelevant rules. Both are applied during network training for early pruning of redundant or irrelevant terms and rules, excluding them from further parameter learning. All steps of the modeling method are presented, and the method is illustrated by an example from the literature.