Assessing the Accuracy of Spatiotemporal Epidemiological Models

  • Authors:
  • James H. Kaufman;Joanna L. Conant;Daniel A. Ford;Wakana Kirihata;Barbara Jones;Judith V. Douglas

  • Affiliations:
  • Healthcare Informatics Research, IBM Almaden Research Center, San Jose, United States of America 95120;College of Medicine, University of Vermont, Vermont, United States of America 05405;Healthcare Informatics Research, IBM Almaden Research Center, San Jose, United States of America 95120;Wakana Kirihata, Columbia University, New York, United States of America New York 10027;Healthcare Informatics Research, IBM Almaden Research Center, San Jose, United States of America 95120;Healthcare Informatics Research, IBM Almaden Research Center, San Jose, United States of America 95120

  • Venue:
  • BioSecure '08 Proceedings of the 2008 International Workshop on Biosurveillance and Biosecurity
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

To demonstrate an approach that allows for the assessment of models and their accuracy, a numerical experiment was designed to generate a "control" data set and treated it as if it were "real" data. The open source spatiotemporal epidemiological modeler (STEM) was used to develop a control scenario depicting the spread of influenza in the state of Vermont; this scenario was then compared to three alternative models using such tools as root mean square differences and phase space analysis. This approach may prove helpful in responding to global pandemics and arriving at necessary policy decisions.