Article Preview
Top1. Introduction
Decisions making process is among the greatest challenges of our developing world. Industries and corporates are often called to make big and essential decisions at each steps varying from the manufacturing phase and all the way up to delivering at the customer’s door-steps. Likewise, in the global supply chain the supplier’s selection is an important key factor to consider for business reliability, as per Van Weele (2014), the selection of suppliers plays a predominant role in the procurement management, as it considerably impacts on the supply chain performance. Supplier’s performance and consistency have greater consequences on the fulfillment of customer’s requirements. The ability and capacity of suppliers to provide required resources is vaguely dependent on various parameters of high interest such as, availability, time management, cost effectiveness, quality insurance, etc., and ensuring the best list of suppliers for the requirements within an extending supplier base can be trickier and challenging over time. As reported by Çebi & Otay (2016); Karsak & Dursun (2015) that, in spite of the difficulties maybe involved in the supplier selection, the process remains a necessity for its substantial influence on the organization operation and profitability. For some of the papers from Reshu & Mittal (2019), Jayaswal et al. (2019), Jaggi et al. (2011), Mittal et al. (2017), proposed supply chain models with various features. Optimization of the process of selecting suppliers is of an exceptional importance for corporates and businesses in the amplitude to enable high efficiency in the process, so by allowing time saving, cost minimization, and a considerable lead time in the delivering of services. A large number of variables, both quantitative and qualitative, can become complex for the selection of appropriate suppliers, according to Adam & Filip (2018). Supplier’s selection of variables is usually dependent on industries specificities and requirements, and as per Deng et al. (2014), companies entertain different strategies and cultures which considerably impact on the variables selection; Kar (2013), also reported on the prioritization approach of selection criteria for supplier. Globally, similarities in the use of certain variables as important basis for supplier’s selection can be observed from the literature survey by Adam & Filip (2018), from which a total number of 29 variables were identified. Newer opportunities are emerging in these recent years and have allowed incredible technology advancements in many areas, such as in AI (Artificial Intelligence) which Jordan & Mitchell (2015) explained the prospect in length. And Machine learning as a sub part of AI (Artificial Intelligence), helps address challenges of our modern world era with its developing techniques, that improve on problem solving, solution finding, optimization of existing process, just as Zhou et al. (2017) reported on its immense opportunities and also challenges; its applications on the optimization of supply chain by Sandhya et al. (2019). Machine learning enables decision makers to act meaningful over variations and changes in the process. For Hurwitz & Kirsch (2018), Machine learning comprises four different learning approaches such as, supervised learning, unsupervised learning, reinforcement learning and deep learning. Quite few recent articles, such by Mohammed et al. (2020), Brewer et al. (2019), Kumar (2019), David (2020), Atefeh (2018), Hayk (2020), Boyce & Mano (2018), attempted to address the similar subject of optimization but using different approaches in determining important set of variables, and as per Zhang et al. (2016), a minimum of sixteen (16) variables can be used with machine learning. Further methodological approaches can be examined for the model of optimization, and since no one solution fits all sort of problem, this paper considers a unique approach in the assumption of 18 common variables of types quantitative and qualitative for the selection of supplier, and an optimization approach based on Logistic regression machine learning technique. The model learns from historical data, and make prediction for best suppliers as to match customer’s requirements. In this study, the subsequent sections include, a theorical analysis in the selection process of supplier, a definition of supplier’s selection framework, a description of the model of optimization based on machine learning algorithm technique, evaluation and discussion on the optimization model, and finally concluding remarks.