A game theory-based decision-making framework for sustainable urban planning: optimizing residential density with the aid of machine learning techniques

Residential density planning is crucial for urban growth, impacting resource use, sustainability, and quality of life. The urban fabric is significantly influenced by the decisions made by planners and stakeholders regarding building footprints. Aligned with the conference’s focus on sustainable solutions, this research introduces a Game-Theoretic Interactive Decision-Making (GTIDM) tool that combines game theory (GT) and machine learning (ML). This framework model’s stakeholder behavior in residential density planning enhances decision-making and promotes sustainable urban development. In complex, competitive, and conflicting decision-making contexts, a game-theoretic framework is used to achieve optimal results by considering all possible scenarios. Three diverse density character zones were selected, including two within and one outside the Colombo Municipal Council (CMC), and subjected to the model's application. Expert validation indicated that while both simultaneous and sequential models replicate realistic data, the simultaneous model is more suitable for determining ideal building density. This study demonstrates the integration of GT and ML as a powerful strategy for individual and group decision-making in urban planning. Accurately calculated payoffs using the GTIDM model, which align with the study’s goals and strategies, are crucial. Rationalizing residential density decisions encourages better stakeholder judgments, thereby promoting sustainable solutions and advancing sustainable urban density practices.