26 Simulation Analytics for Social and Behavioral Modeling

Samarth Swarup Achla Marathe Madhav V. Marathe and Christopher L. Barrett

Biocomplexity Institute & Initiative, University of Virginia, Charlottesville, VA 22904, USA

Introduction

Computational simulations are increasingly being used to study social systems, 1 in order to model a range of phenomena, including infectious disease epidemics, natural‐ and human‐initiated disasters, economic self‐organization, online social behavior, and more. These phenomena are behaviorally driven and exhibit coupling between multiple systems and multiple spatiotemporal scales and also offer many opportunities for interventions. These modeling and simulation efforts are giving rise to interesting new questions that need to be addressed through new methods that combine data analytics and simulation science, which we refer to as simulation analytics.

Simulation‐based approaches are needed for these complex problems because policy and planning require understanding hypothetical (counterfactual) scenarios, answering what‐if and under‐what‐conditions questions (Davis and O'Mahony 2019), and discovering interventions. High‐resolution and data‐driven simulations of adequate representational complexity provide a natural way of addressing these requirements. For instance, if during an infectious disease epidemic, an epidemiologist wishes to understand the potential consequences of closing a particular school for a certain number of ...

Get Social-Behavioral Modeling for Complex Systems now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.