The Role of Intelligent Technologies in Advancing Operations Research Models and Statistical Methods: A Review

Abstract

This article reviews the evolving role of smart technologies in advancing operations research models and statistical methods. The article begins by asserting that operations research and statistics have developed relatively separately despite their natural complementarity, as statistical methods provide the data and standards needed for operations research models. The article identifies three main categories of smart technologies: Artificial Intelligence and Machine Learning (sub-categories: supervised learning, unsupervised learning, and reinforcement learning), data infrastructure and big data analytics (including data processing and feature engineering), and automation, simulation, and digital twining. The article reviews the synergies between these technologies and process research modeling through three main areas: model formulation and solution techniques, by integrating machine learning predictions, surrogate modeling, and designing hybrid structures, data-driven calibration and validation: using data to derive model parameters and estimating uncertainty, instantaneous optimization and adaptive decision-making: through rapid re-optimization and multi-stage optimization. The article also discusses the enhancement of statistical methods through intelligent technologies, especially in AI-powered Bayesian and probabilistic modeling, non-parametric methods, and powerful methods in intelligent data environments, causal reasoning, and experimental design with learning-enhanced tools. The article concludes with applications across multiple sectors, including supply chains and logistics, healthcare and epidemiology, energy systems and sustainability, and finance and risk management, emphasizing that the integration of smart technologies with operations research and statistics opens up new avenues for improving decision-making in complex and data-rich environments.