statnet : Introduction| Installation | Repository | Resources | FAQ | Citing statnet
NME workshop : Location | Lectures | Tutorials | Labs | Discussion
Network Modeling for Epidemics (NME)
6 day Summer Workshop
University of Washington, Seattle WA
July 8-13, 2013
Mathematical modeling plays a growing role in infectious disease epidemiology, for studying the dynamics of pathogen invasion and persistence, understanding the sources of disease disparities among populations, and predicting the impact of interventions. Deterministic compartmental models (based on ordinary differential equations) have been the traditional basis for this work during the past two decades. Recent advances in statistical theory and methods have given rise to a new class of stochastic “agent-based” network models, however, and these models are more appropriate for small scale assessments, where the effects of chance lead to wide variation in potential outcomes, or when infection is spread by a small number of highly structured contacts, as with HIV and other STIs.
This course starts with a brief review of deterministic SI, SIR and SIS models, and then provides an introduction to the new stochastic network models for epidemics, with a focus on empirically based modeling of HIV transmission and control. It will be a “hands-on” course, with integrated lectures and computer lab sessions devoted to programming. Labs will be based on the user-friendly computational tools available in the package EpiModel (a free package that uses the R programming language). EpiModel can be used for both deterministic and stochastic network modeling, so the user only needs to learn one package. EpiModel seamlessly integrates the functionality of the statnet software to handle the network modeling. This allows the user to estimate generative network models from empirical data, and then use those models to simulate dynamic transmission networks with the observed properties, analyze the results (using all the functionality of R packages) and visualize the results as network movies. The course outline for the week is:
- Monday July 8th
- Introduction to epidemic modeling
Deterministic and stochastic models for epidemics
Lab: Programming simple deterministic and stochastic SI, SIS and SIR models with EpiModel
- Tuesday July 9th
- Classical network analysis
Introduction to Exponential Random Graph Models (ERGMs) for networks
Lab: Using statnet for static network modeling and visualization
- Wednesday July 10th
- ERGMs: Separable temporal ERGMs (STERGMs) for dynamic networks
Egocentric network data in ERGMs and STERGMs
Lab: Using statnet for dynamic network modeling and visualization
- Thursday July 11th
- Epidemic models on networks: Independent network and epidemic processes
Lab: Using the EpiModel epidemic modeling tools for dynamic networks with a fixed population
- Friday July 12th
- Epidemic models on networks: Dependent network and epidemic processes
Lab: EpiModel and dynamic network models with changing population size/composition
- Saturday July 13th
- Open, to discuss projects and research ideas of interest to students
Saturday attendance is optional
Participants will develop the basic tools and knowledge needed to specify and program models for their own research projects.
LOGISTICS
Dates: Monday, July 8 – Saturday, July 13, 2013
Times: 10 am (sharp) – 4 pm
Location: Waterfront Activity Center Meeting room, Main campus of the University of Washington in Seattle. Directions.
Accomodations at UW: McMahon Hall, Main campus of the University of Washington in Seattle.
Instructors: Professors Martina Morris and Steven Goodreau
Costs: Registration is $250 On-campus housing will be available for about $70/night. A small number of fellowships will be provided to predoctoral students and/or international (non-US/Canada) scholars
Target audience: Researchers and students in any field with an interest in epidemic modeling
Course webpage: https://statnet.csde.washington.edu/trac/wiki/NME2013
Prior experience: Basic familiarity with R is required (but see below). Previous modeling experience (broadly defined) is recommended.
Those with general statistical/modeling skills but no knowledge of R may apply now, and obtain familiarity with R in the interim, via
- http://www.cran.r-project.org/manuals.html or
- http://csde.washington.edu/courses/statcore/RIntro/
- http://www.statmethods.net/
You may also want to look through some of the tutorials for the statnet package.
Questions? Email us: goodreau at uw.edu, morrism at uw.edu
Steven M. Goodreau / Assoc. Professor / Dept. of Anthropology / University of Washington http://faculty.washington.edu/goodreau
Martina Morris / Professor of Sociology and Statistics / University of Washington http://faculty.washington.edu/morrism/
Attachments (2)
- NME 2013 Informational Flyer.docx (23.8 KB) - added by goodreau 4 years ago.
- NME 2013 Informational Flyer.2.docx (23.8 KB) - added by lxwang 4 years ago.
Download all attachments as: .zip