Published Paper Details:

The application of machine learning (ML) into decision support systems (DSS) has produced high levels of accuracy, efficiency and adaptivity in several sectors. This review charts the evolution of ML in DSS, illustrating important milestones, applicable techniques, and the challenges encountered for these systems as they matured. This paper explores the most popular models related to ML used in DSS including decision trees, SVMs, NNs and ensemble methods and their relative advantages and disadvantages for their application in practical settings. Discussed are basic questions such as model interpretability, data quality, scalability, and real-time processing, as well as the related development of solutions that are rooted in explainable AI (XAI), data preprocessing, as well as deployment of edge and cloud computing technologies. This analysis sheds light on main achievements and the existing challenges in real world application of ML in DSS and highlights important research questions in order to further develop the practical and responsible application of such systems.