A Self-Adjusting Decision Support System for Project Portfolio Prioritization using Integrated AHP and Reinforcement Learning

This is the NetLogo Agent-Based Model (ABM) source code used for the research study: 'A Reinforcement Learning Framework for Adaptive Decision-Making in Project Portfolios'. Purpose and Methodology: The model implements an adaptive decision-making framework for optimizing strategic project prioritization and achieving Project Portfolio Management (PPM) resilience under persistent environmental uncertainty. The framework employs a hybrid approach integrating the Analytical Hierarchy Process (AHP) for establishing Key Performance Indicator (KPI) weightings and a Q-learning algorithm (governed by an ε-greedy policy) for sequential portfolio selection. Findings & Application: The simulation, demonstrated within this code, shows that the RL-based approach significantly improves the strategic alignment and operational adaptability of the project portfolio compared to traditional methods. The model provides a novel mechanism to connect short-term project actions to long-term strategic organizational objectives. This code serves as the primary software data for the research article and can be used by researchers for replication, validation, and future development of RL-AHP integrated systems in complex organizational contexts.