Substantial progress has been made over the past 10 years in the methods needed to model epidemic dynamics on stochastic networks. True network modeling, in contrast to agent-based modeling, focuses on replicating observable network characteristics, in addition to individual behavior. The methodology for this has roots in the social network literature, and, more recently, in statistics. The new statistical methods are based on Exponential Random Graph Models (ERGMs), and are designed to estimate models from complete or egocentrically sampled network data. The estimated models can be used to simulate dynamic networks that have the properties observed in the empirical data, and thus establish a foundation for studying the spread of epidemics on networks. This talk will introduce the ERGM framework and the extensions most relevant for epidemic modeling, with examples of current applied modeling projects.
About the Speaker
Dr. Morris holds a joint appointment as professor in the departments of Sociology and Statistics at the University of Washington, and is the founding director of the Sociobehavioral and Prevention Research Core in the UW CFAR. Her work focuses on the development of statistical methodology for network analysis, applications to modeling epidemics on networks, and developing local models for HIV prevention planning.