Data Mining Classification Techniques to Improve Decision-Making Processes

‎This study aims to enhance decision-making processes in dynamic organizational environments by integrating Artificial Intelligence (AI) techniques. It explores AI's potential to improve efficiency, accuracy, and predictability while addressing the ethical concerns, biases, and risks of over-reliance that accompany its implementation. We introduce a novel, domain-independent iterative methodology that leverages data mining classification techniques to analyze large datasets from information systems. This approach identifies specific situations where existing decision-making strategies yield suboptimal results and uses these insights to refine and improve the strategies employed. The methodology has proven effective in augmenting feedback control strategies by employing an inductive machine learning algorithm to uncover areas for enhancement. Our results indicate that initial strategies can be upgraded to achieve comparable levels of cost efficiency, even when accounting for the costs associated with evaluating potential strategies. This study recognizes limitations related to the variability of expert opinions in scenarios characterized by numerous components, which may constrain the optimization scope of the learning system. Future research should consider the application of advanced optimization techniques, such as genetic algorithms, to better establish optimal conditions for improvement. The findings suggest that this methodology can be readily applied within simulation frameworks, allowing organizations to assess changes in control strategies effectively. This facilitates informed resource allocation and improves operational efficiencies across diverse settings. As the adoption of AI in decision-making processes increases, attention to ethical considerations becomes crucial. Our methodology promotes transparency and addresses potential biases, fostering responsible AI use that mitigates negative societal impacts.