Environmental policy in the context of complex systems: Statistical sensitivity testing and policy optimization using agent-based models
, 1900
This paper introduces a machine-learning–assisted framework for optimizing environmental policy in complex agent-based models. By combining statistical sensitivity testing with Bayesian optimization, it enables efficient policy search despite costly simulations. Applied to an extended Sugarscape model with pollution, the approach identifies interpretable policies and reveals key tradeoffs between welfare, inequality, and survival.
Recommended citation: D. Munson, A. Dey, and S. Mak (2026+). Environmental policy in the context of complex systems: Statistical sensitivity testing and policy optimization using agent-based models.
