Development of a complex method for finding a solution for neuro-fuzzy expert systems

Abstract

Artificial intelligence has become the backbone of modern decision sup port systems. This is why a complex method for finding solutions for neuro fuzzy expert systems has been deve loped. The proposed complex meth od is based on a mathematical model for the analysis of the operational si tuation. The model makes it pos sible to determine the parameters of the analysis of the operational sit uation, their influence on the qual ity of assessment of the operation al situation and to determine their number with units of measurement. An increase in the efficiency of infor mation processing (error reduction) of the assessment is achieved by the use of evolving neurofuzzy artificial neural networks. Training of evolv ing neurofuzzy artificial neural net works is carried out by training not only synaptic weights of the artifi cial neural network, the type, parame ters of the membership function, but also by applying the procedure for reducing the dimension of the feature space. The efficiency of information processing is also achieved by train ing the architecture of artificial neu ral networks; accounting for the type of uncertainty in the information to be assessed; work with both clear and fuzzy data. We achieved a reduction in computational complexity while mak ing decisions; the absence of errors in training artificial neural networks as a result of processing information entering the input of artificial neural networks. The analysis of the opera tional situation as a whole occurs due to the improved clustering procedure, which allows working with both static and dynamic data. The proposed com plex method was tested on the example of assessing the state of the operatio nal situation. The mentioned example showed an increase in assessment effi ciency at the level of 20-25 % in terms of information processing efficiency