Pros and Cons of ABMs for Epidemic Modeling-Virtual Workshop

Registration is now closed

June 21-22, 2021

Agent-Based Models (ABMs) are widely used in epidemiology, with notable applications to Ebola, malaria, and COVID-19. However, the statistical properties of these models are underdeveloped, which makes quantified statements of uncertainty about forecasts difficult. Additionally, there is no standard theory for how to fit parameters in such models, nor how to assess model adequacy. On the other hand, the flexibility and face-validity of ABMs is a powerful tool for conducting “what-if” simulations that can guide public policy. This workshop brings together experts to discuss the capabilities, limitations, and potential pitfalls of ABM use in studying disease spread.

Confirmed speakers/panelists:

Paul Birrell, National Infection Service, UK
Georgiy Bobashev, RTI
Sara Del Valle, Los Alamos National Laboratory
Jonathan Fintzi, NIH
M. Elizabeth Halloran, Fred Hutchinson Cancer Research Center/University of Washington
Mevin Hooten, Colorado State University
Nianqiao (Phyllis) Ju, Harvard University & Purdue University
Chris Krapu, Oak Ridge National Laboratory
Lucy M. Li, The Public Health Company
Jessica Notestine, NCSU
Bruce Rogers, formerly Duke
Toryn Schafer, Cornell University
Simon Spencer, University of Warwick
Anuj Srivastava, Florida State University

Schedule

Speaker Titles and Abstracts

Recorded Talks Monday June 21st

June 21st
10 am – 2pm EDT
(3pm – 7pm in London,  8am – 12pm in Colorado & New Mexico, and 7am – 11am in Seattle)

9:50 – 10am Opening remarks
10:00 – 10:30am Paul Birrell, National Infection Service, UK

“Real-time Challenges in Modelling a Real-life Pandemic”

10:30 – 11:00am Georgiy Bobashev, RTI

“Hybrid System Dynamics and Agent-Based Models”

11:00 – 11:30pm Chris Krapu, Oak Ridge National Laboratory

“Making the Most of Recent Advances in Statistical Programming for ABM Emulation”

11:30 – 12:00pm Bruce Rogers, formerly Duke

“How Much Testing is Enough?”

12:00 – 12:15pm Break
12:15 – 12:45pm Anuj Srivastava, Florida State University

“Agent-Level Pandemic Simulation (ALPS) for Quantitative Assessment of Lockdown Measures”

12:45 – 1:15pm Toryn Schafer, Cornell University

“Inverse Reinforcement Learning for Agent-Based Models”

1:15pm – 1:45pm M. Elizabeth Halloran, Fred Hutchinson Cancer Research Center/University of Washington

“Simulations for Designing and Interpreting Vaccine Studies”

Recorded Talks June 22nd – AM

Recorded Talks June 22nd – PM

June 22nd
10am-2pm EDT
(3pm – 7pm in London,  8am – 12pm in Colorado & New Mexico, and 7am – 11am in Seattle)

10:00 – 10:30am Simon Spencer, University of Warwick

“Scalable Bayesian Inference and Model Selection for Individual-based Epidemic Models”

10:30 – 11:00am Nianqiao (Phyllis) Ju, Harvard University & Purdue University

“Scalable Inference in Agent-based Epidemic Models using Sequential Monte Carlo”

11:00 – 11:30am Jessica Notestine, NCSU

“Sensitivity and Active Subspace Analysis for Agent-Based Models”

11:30 – 11:45am Break
11:45 – 12:15pm Sara Del Valle, Los Alamos National Laboratory

“Modeling COVID-19 with Agent-based Simulations: Challenges & Opportunities”

12:15 – 12:45pm David Banks, Duke University / SAMSI

“Inference Without Likelihood”

12:45 – 1:25pm Panel discussion
1:25 – 1:30pm Closing remarks

Questions: dsbs@samsi.info