The use of confidence measures to enhance combination strategies in multi-network neuro-fuzzy systems

It is well known that substantial improvements can be obtained in difficult pattern recognition problems by combining or integrating the outputs of mu ltiple neural classifiers. This paper analyses the performance of some comb ination schemes applied to a multi-hybrid neural system which is composed o f neural and fuzzy neural networks. Essentially, the combination methods em ploy different ways to extract valuable information from the output of the experts through the use of confidence (weights) measures of the ensemble me mbers to each class. An empirical evaluation in a handwritten numeral. reco gnition task is used to investigate the performance of the presented method s in comparison with some existing combination methods.