Workshop on Verification and Validation of Epidemiological Models
Washington D.C., USA
Date: December 14-16, 2012
http://vvm2012.myvnc.com/workshop/

Scope:

Epidemiological models are designed for examining the process of an epidemic (disease spread) in population. The results of these models are used as inputs to critical decisions on the prevention, containment, and countermeasures (e.g., immunization schedules) for the disease under study. Given the criticality of these decisions, it is vital that the information upon which they are built is trustworthy. Verification and validation (V&V) are fundamental to establishing trust in simulation models and their results.

In general, epidemiological models belong to a class of software sometimes referred to as “non-testable programs”. Most agent-based models, including agent based epidemiological models fall into this category, which, as described by Davis and Weyuker, are “programs which were written in order to determine the answer in the first place. There would be no need to write such programs if the correct answer were known.” Since there are no oracles for these programs, it is generally impossible to know a priori what the correct or expected output should be for a given set of inputs. There are many criticisms about using ABM to study complex systems and disease spread because of the lack of maturity and widely accepted model validation techniques. Because of the heterogeneity of the agents and the possibility of new patterns of macro behavior emerging as a result of agent interactions at the micro level, model validation in agent-based complex social systems is different from the traditional validation. Independent intelligence of agents, the large number of concurrent and non-trivial interactions between those members, varying rates of learning and what is learned, and other factors make the validation of ABM difficult. Most current V&V methods cannot be used verify or validate vital aspects of agent-based epidemiological models, nor can they provide a measure of confidence that can be reasonably placed in the model results. V&V can lend confidence to the use of models to inform critical decisions. The importance of simulation model V&V increases as their results are used in increasingly critical decisions.

Topics:

The aim of this workshop is to share research and experience in the V&V of epidemiological models in order to prepare an overview of current status of the field and to discuss the techniques and issues in this domain. This workshop will provide a stage for scientists to present their research in relevant verification and validation techniques, complex system case studies, and model comparison approaches. We welcome contributions in, but not limited to, the following areas:

• Advances in sensitivity analysis
• Agent-based model V&V
• Comparison of differential equation and agent-based disease spread models
• Differential equation model validation
• Disease spread model V&V
• Emergent behavior V&V
• Empirical validation
• Model checking and model-based testing
• Model parameter exploration
• Population model V&V
• Social network and social interaction model validation
• Statistical and mathematical V&V
• Sub-model interaction analysis
• Transportation model validation

Important Dates:

Paper submission deadline: September 15, 2012
Author Notification: October 15, 2012
Final Manuscript Due: November 1, 2012
Conference Date: December 14-16, 2012

Paper Submission:

Each submission should not exceed six (6) pages including title, authors, tables, figures and references. The font size should be no smaller than 11 point. Submission of a paper represents a commitment that, should the paper be accepted, at least one of the authors will register for and attend the conference to present the work. Please email your contribution to Dr. Xiaohui Cui at xcui@nyit.edu and CC Dr. Laura Pullum at pullumll@ornl.gov

Program Chairs:

Dr. Laura Pullum, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
Dr. Xiaohui Cui, New York Institute of Technology & Oak Ridge National Laboratory, Manhattan, NY 10023, USA

Program Committee:

Dr. Mike Hinchey, Irish Software Engineering Research Center, University of Limerick, Ireland
Dr. Chris Rouff, Near Infinity LLC, USA
Dr. Dylan George, United States Department of Defense, USA
Dr. Charles Pavia, New York College of Osteopathic Medicine, New York Institute of Technology, USA