Machine Learning Assisted Multi-Criteria Decision-Making Approaches for Site Selection: A Systematic Review

Choosing a site for a business is a crucial decision that can have lasting effects on sustainable growth and competitiveness. Location selection involves a multi-dimensional decision-making process that influences long-term performance, cost efficiency, and operational success. Using Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and the Context-Intervention-Mechanism-Outcome (CIMO) framework, we systematically reviewed 189 peer-reviewed studies (2015-2025) to understand how different methods are applied across domains, contexts, and criteria in site-selection problems. Results show a clear shift from classical MCDM models to hybrid (MCDM+ML) frameworks. These hybrids combine transparency and interpretability with predictive strength. Building on this evidence, our review provides the first cross-domain taxonomy tracing the methodological evolution from classical MCDM to hybrid MCDM+ML approaches. The challenges include the absence of benchmark datasets, the lack of methodological standards, and the difficulty of cross-domain validation. Hybrid frameworks balance transparency, predictive capability, and stakeholder trust, and they are adaptable across domains with minimal model changes. This review establishes a roadmap for benchmarking, reproducibility, and validation, positioning hybrid MCDM+ML frameworks as standardized, trustworthy tools for high-stakes site selection.