A Study of Paraconsistent Artificial Neural Cell of Learning Applied as PAL2v Filter

Abstract

The Paraconsistent Annotated Logic (PAL) is one type of non-classical logics that, differently than classical logic, allows the processing of contradictory signals in its theoretical structure. The Paraconsistent Artificial Neural Cell (PANcell) is the basic block of a set of algorithms that utilizes the interpretation of the lattice of the Paraconsistent Annotated Logic with annotation of two values (PAL2v). The Paraconsistent Artificial Neural Cell of Learning (LPANcell) is a type of PANcell that uses the output as a feedback to the input of the unfavorable degree of evidence in time. The resulting behavior is to learn any real value within a normalized closed range (interval values [0, 1]). This cell can be used in signal analysis, processing, average estimator and as a filter, called PAL2vFilter. The proposed article presents the results of the analysis of a PAL2vFilter, when using the simplified resulting degree of evidence - uE or the resulting degree of real evidence - uRE (when contradiction effects are extracted) as output.