Another class of fuzzy connectives in fuzzy neural networks

In this paper we have attempted to understand another class of the T-norms and T-conorms, mainly those which are generated by the simple, monotonic, continuous, non-conditional functions. Fuzzy connectives based on those T-operators play an important role in encoding and decoding of fuzzy neural networks. Learning aspects of neural networks can be related to the relation matrix in fuzzy theory. Similarly, the recall procedure of fuzzy inferencing can also be mapped to the reasoning in neural networks. We propose a new additive-product connective generator different from the ones known in fuzzy literature. The exponential norms generated from the proposed connective generator can give rise to various triangular operators with different strength due to the variation of the control parameters, and that affects the learning and the reasoning behavior of a fuzzy neural network.