Population Synthesis

Session chairs:

Gary Polhill, The James Hutton Institute, Aberdeen, UK

Mario Paolucci, National Research Council, Institute for Research on Population and Social Policies, Rome, Italy

Rocco Paolillo, National Research Council, Institute for Research on Population and Social Policies, Rome, Italy

Clémentine Cottineau, TU Delft, The Netherlands

Tatiana Filatova, TU Delft, The Netherlands

Description:

This track pertains to the challenges of generating synthetic populations for agent-based models. This array of techniques can support modellers against issues such as privacy and data protection when using data archives of micro-level data, and related issues from integrating data across multiple data sets. While population synthesis is an ongoing area of research in the microsimulation community, agent-based models pose additional challenges stemming from the need to include mechanisms of interaction: where to locate agents, what their daily routines are, with which other agents they interact, and how they make their decisions. Plausible and/or principled means for creating synthetic populations bringing together these various features in ways that are representative of the real underlying population is necessary to provide deeper insights into policy scenarios taking full advantage of the heterogeneous characteristics of agents in the dynamics of agent-based simulations. We welcome submissions on all aspects of creating synthetic populations for empirical agent-based models.