Title: Stochastic computer model calibration using sequential monte carlo
Speaker: Arindam Fadikar
Time: 2:30PM
Abstract: Dynamic network simulation models can evolve a very large system of agents over a network that changes with time.  Such models are often used to simulate epidemics or transportation, typically producing random trajectories, even when model parameters and initial conditions are identical. This introduces a number of challenges in designing ensembles of model runs for sensitivity analysis and computer model calibration. Gaussian process meta modelling combining observations and experimental data is a popular and well-known technique for uncertainty quantification task, but it requires significant amount of computing power when model dimension increases; it also has difficulty with random output. The present work focuses on a novel model calibration technique based on a sequential monte-carlo approach, to an epidemic simulator, seeking to forecast the epidemic’s behavior by geographic region. Based on a simple and interpretable importance sampling approach, this method sequentially selects plausible model parameter settings from a pool of design points, enabling exploration of the posterior and direct description of prediction uncertainty.

Title: Understanding Diffusion Processes on Networks using Network Reliability
Speaker:  Yihui (Ray) Ren
Time: 3:15PM

How to develop a real-time surveillance system for epidemics (or other network stochastic processes you can think of)? Which people do you choose to ask the question “are you sick”? How many people do you want to pick so that you are certain your estimation is accurate enough? Given a set of survey results, can you say anything about the epidemics, i.e. the transmissibility? How certain are you about your estimation? These are all interesting questions, but I have no answer! Come to the talk with pen and paper. We can try to solve them together. :D 

Webex details:
Meeting number: 643 164 789
Meeting link:
Host key: 749673
Audio connection:
+1-855-749-4750 US TOLL FREE
+1-415-655-0001 US TOLL
Show toll-free dialing restrictions
Access code: 643 164 789

Event Contact:
Srini Venkat