Data-Driven Approaches for Project Portfolio Selection Using Multi-Criteria Decision-Making (MCDM) Methods

Abstract

Project Portfolio Selection (PPS) is a critical decision-making process in project-oriented organizations, requiring evaluation across multiple conflicting objectives, including cost, risk, sustainability, and strategic alignment. Traditional Multi-Criteria Decision-Making (MCDM) methods, while effective, often rely on subjective weight assignments from experts, which can introduce bias. Recent advances in data-driven decision-making enable the integration of machine learning, big data analytics, and automatic weight-generation methods with MCDM, enhancing objectivity and robustness. This study presents a comprehensive framework for data-driven PPS using MCDM methods, supported by a literature review (2019–2025), a mathematical formulation, and a numerical illustration. Results indicate that data-driven weighting enhances portfolio efficiency, mitigates uncertainty, and aligns selection with organizational strategy. Identified research gaps suggest the need for large-scale empirical validation, hybrid machine learning–MCDM models, and cross-industry applications.