Modeling and Optimizing Agent-Based Model of Conflict-Induced Forced Migration

Conflict-induced forced migration leads to large-scale displacement of populations, causing social and economic disruption. The recent Russian invasion of Ukraine serves as a stark example, resulting in millions of internally and internationally displaced people. Aiding them requires an understanding of the various dynamics of forced migration, including pre-migration intentions and postmigration outcomes such as return migration. However, existing computational approaches in the literature lack a cohesive framework that integrates these multi-dimensional aspects of forced migration using social and behavioral theories as foundational elements. My dissertation focuses on developing an a) end-to-end generalized agent-based model (ABM) of forced migration that captures these multi-phased dynamics, b) uses social and behavioral theory in modeling agent behavior, and c) applies various optimization and non-differentiable techniques to jointly optimize these multi-dimensional dynamics of forced-migration. I also perform various case studies that highlight the policy relevance of the model.