Methodology for comparing agent-based models of land-use decisions

The focus of the research is mainly methodological. The goal is to develop a framework for comparing computational, agent-based models for land-use decision making. The framework will allow studying spatial patterns emerging from different distributions of agent characteristics, learning and communications schemes, initial spatial configurations, and varying spatial suitabilities. The framework will also facilitate the assessment of model complexity, thus enabling a rigorous practice of model comparison and selection.