SUSTAINABLE SUPPLIER SELECTION THROUGH MACHINE LEARNING AND MCDM METHODS

Abstract

One of most important task for any department of a company, preferably purchasing department is to select the best suppliers since selection of supplier most affect the performance of whole supply chain and hence the value of supply chain. Selection of right supplier for a company’s manager can be a tedious process since it comprises of many number of variables and has both qualitative and quantitative variables which needs to be consider for it. The purpose of this study is to give the more weightage to the Environmental criteria by considering more number of criteria which are related to Environment. This study aims to implement hybrid mode of study in which first part of the problem Machine learning will be applied over the different alternatives and criteria, Machine learning algorithm will reduce the large number of suppliers into a bunch of best suppliers and will also reduce the time as compared to conventional Multi criteria decision making(MCDM) methods. In the second part of problem MCDM methods has been applied to find final best suppliers. As a Methodology in first half of the problem Clustering algorithm has been applied to make a cluster of best supplier and in second half of the problem TOPSIS (a MCDM method) has been used to find best supplier from the best cluster. This study incorporated 34 parameters which are imperative for supplier selection, out of these 34 variables major focus was on the Environmental criteria. Apart from these a unique combination of the criteria has been selected in which Technological criteria has been merged with the sustainable criteria.