Towards Paraconsistent Neuroscience: a Review Paper on Some Applications of PANN

In this expository work, we sketch some applications of a new theory of Artificial Neural Network - ANN, based on a paraconsistent annotated evidential logic Eτ. Such theory, called Paraconsistent Artificial Neural Network - PANN - has as characteristics the capability of manipulating uncertain, inconsistent and paracomplete concepts directly without trivialization. PANN differs from other usual ANNs which are based on classical logic or in some of its extensions. Some aspects such as the capability of adaptation, velocity processing, and other useful characteristics make the PANN a promising theory. Although there are some essential non-classical approaches for reasoning (v.g Fuzzy Set Theory) in this paper, we discuss how PANN can be a basis for a reasoning model for the human brain. As the logic Eτ encompasses the classical logic and Fuzzy reasoning, we believe that PANN can be the basis for a new model for the computational scope of the Neuroscience.