An Innovative Technique for Intelligent Decision Making: Smart TOPSIS using Naïve Bayes Classification Algorithm

Abstract

This research work is based on the development of an intelligent Multi-Criteria Decision-Making Method. Research is mainly focused towards conversion of TOPSIS from its conventional form to an intelligent form. We named this intelligent variation as SMART TOPSIS. Smartness of the TOPSIS is due to implementation of supervised learning Naïve Bayes classification algorithm which allows decision makers to compute the weights of each criterion probabilistically, especially by Bayes Theorem. So, here our basic thrust has been given to implement Naïve Bayes classification for weights of the various criteria. TOPSIS has been chosen for this task because it is simple, rationale and comprehensive. Efficiency of computation in TOPSIS is very high and the mathematics behind the measurement of relative performance for each alternative is simple. The knowledge base of SMART TOPSIS is dynamic in nature because Bayesian classification for learning the weights of the criteria of a decision-making model under TOPSIS is considered dynamically. The paper discusses the machine learning details towards dynamic Bayesian classification for selection of the best alternatives among many.