Neal Altman

BioWar: scalable agent-based model of bioattacks

While structured by social and institutional networks, disease outbreaks are modulated by physical, economical, technological, communication, health, and governmental infrastructures. To systematically reason about the nature of outbreaks, the potential outcomes of media, prophylaxis, and vaccination campaigns, and the relative value of various early warning devices, social context, and infrastructure, must be considered. Numerical models provide a cost-effective ethical system for reasoning about such events. BioWar, a scalable citywide multiagent network numerical model, is described in this paper. BioWar simulates individuals as agents who are embedded in social, health, and professional networks and tracks the incidence of background and maliciously introduced diseases. In addition to epidemiology, BioWar simulates health-care-seeking behaviors, absenteeism patterns, and pharmaceutical purchases, information useful for syndromic and behavioral surveillance algorithms.

Multi-Agent Modeling of Biological and Chemical Threats

When pandemics, chemical spills, and bio-warfare attacks occur cities must respond quickly to mitigate loss of life. Which interventions should be used? How can we assess intervention policies for novel and low frequency events? Reasoning about such events is difficult for people due to the high level of complexity and the multitude of interacting factors. Computational models, however, are a particularly useful tool for reasoning about complex systems. In this paper, we describe a multi-agent dynamic-network model and demonstrate its use for policy assessment. BioWar is a city-level multi-agent dynamic-network model of the impact of epidemiological events on a city’s population. Herein, we describe BioWar and then use it to examine the impact of school closures and quarantine on the spread and impact of pandemic influenza. Key aspects of the model include utilization of census data to set population characteristics, imputed social networks among agents, and flexible disease modeling at the symptom level. This research demonstrates that high-fidelity models can be effectively used to assess policies.

A Model of Biological Attacks on a Realistic Population

The capability to assess the impacts of large-scale biological attacks and the efficacy of containment policies is critical and requires knowledge-intensive reasoning about social response and disease transmission within a complex social system. There is a close linkage among social networks, transportation networks, disease spread, and early detection. Spatial dimensions related to public gathering places such as hospitals, nursing homes, and restaurants, can play a major role in epidemics [Klovdahl et. al. 2001]. Like natural epidemics, bioterrorist attacks unfold within spatially defined, complex social systems, and the societal and networked response can have profound effects on their outcome. This paper focuses on bioterrorist attacks, but the model has been applied to emergent and familiar diseases as well.

300 Cities Virtual Experiment

This report provides an overview of the preparations required for the virtual experiment we will conduct for the IRS as part of the 300 cities subproject. We briefly describe the tax gap and taxpayer issues, our revised approach, the Construct framework and the models developed for the multi-agent simulation. Where appropriate, we provide references to other technical reports that describe in more detail the models for intentional and inadvertent taxpayer errors, and paid preparers. We also briefly describe how we populate Construct with agents representative of the populations of U.S. cities by sampling from U.S. census data, deriving relevant taxpayer issues for each agent, generating empirically reasonable social networks for each agent, and building Construct input decks automatically. The generation of social networks based on the socio-demographic attributes of individuals found in census data is an advance worthy of the more detailed description found yet another technical report. We then briefly describe the design and anticipated analysis of the 300 cities virtual experiment. We conclude with a brief reference to the SmartCard application that will be used to deliver the results of the virtual experiment along with socio-demographic information and taxpayer issues for each of the cities. Details of the implementation of the SmartCard can be found in the referenced report.