Introducing the Discriminative Paraconsistent Machine (DPM)

Abstract

This paper introduces a new tool for pattern recognition. Called the Discriminative Paraconsistent Machine (DPM), it is based on a supervised discriminative model training that incorporates paraconsistency criteria and allows an intelligent treatment of contradictions and uncertainties. DPMs can be applied to solve problems in many fields of science, using the tests and discussions presented here, which demonstrate their efficacy and usefulness. Major difficulties and challenges that were overcome consisted basically in establishing the proper model with which to represent the concept of paraconsistency.