A Bayesian Analysis of Migration Pathways using Chain Event Graphs of Agent Based Models

Abstract

Agent-Based Models (ABMs) are often used to model migration and are increasingly used to simulate individual migrant decision-making and unfolding events through a sequence of heuristic if-then rules. However, ABMs lack the methods to embed more principled strategies of performing inference to estimate and validate the models, both of which are of significant importance for real-world case studies. Chain Event Graphs (CEGs) can fill this need: they can be used to provide a Bayesian framework which represents an ABM accurately. Through the use of the CEG, we illustrate how to transform an elicited ABM into a Bayesian framework and outline the benefits of this approach.

Bayesian Model Averaging of Chain Event Graphs for Robust Explanatory Modelling

“…They can provide compact representations of complex independence statements, with each different CEG providing a different explanation of the underlying process. CEGs have been successfully applied in various domains including educational studies [Freeman and Smith, 2011], policing [Bunnin and Smith, 2021], public health [Shenvi and Smith, 2019] and migration studies [Strong et al, 2021].…”

“…Other than the ability to represent context-specific independence statements, CEGs differ from BNs by being able to represent domains where processes can unfold in diverse ways. These CEGs are called nonstratified CEGs and examples of their use can be seen in Shenvi and Smith [2019] and Strong et al [2021].…”