Hybrid Intelligence Model on the Second Generation Neural network

Abstract

In latest beyond, it has been visualised in various neural computing applications that the hybridisation of computational intelligence techniques yields better results over traditional methods. This proposed research introduces a new composite version (hybrid) computational model in both real and complex domain that is a unique fusion of revised evolutionary fuzzy clustering along with correlated neural networks. The whole methodology is based on two type of working framework as fuzzy assessment framework and neural categoriser framework. In first framework, suggest styles are disbursed as according to the quantity of clusters on the basis of learning strategy like revised evolutionary fuzzy clustering for generalisation of various evaluated cluster patterns that is totally responsible for selection process of network structure. The second framework of the proposed strategy is to fully based on the correlated neural network evolving the processes of training and subsequent generalisation. The existing real domain analysis is based on first generation neural network (RVNN) and for the analysis of proposed model in second generation neural network (CVNN), Hilbert transformation is used that is available in MATLAB for converting the real datasets in complex form.