Artificial intelligence and machine learning methods for model discovery in the social sciences

Many theories have been proposed for how problematic social phenomena - such as segregation, inequality, and clustering of unhealthy behaviours - emerge and change over time. But which theories are useful for understanding these phenomena and devising effective interventions to change them? This project will develop computationally efficient model discovery methods to enable a search for explanatory multi-level agent-based models that can be calibrated to - and validated against - such empirical phenomena.