Probabilistic Network-Based Epidemic Modelling for Scalable Public Health Simulation

A novel probabilistic approach to epidemic modelling

Objective

• Develop a new probabilistic approach to modelling the instantaneous complex network of social interactions., Develop computational strategies that bypass the need for information transfer between CPU and GPUs., Develop massively parallelizable codes that are amenable for GPU computing., Integrate these developments with existing agent based network models.

Description

• The recent pandemic has highlighted the importance of accurate mathematical models for predicting the propagation of a disease through the population, such that regulatory bodies can adopt appropriate mitigating measures., The focus of this study is on the development of a new probabilistic approach in modelling the complex network of social interactions. This is a novel idea that has never been attempted in network science literature.

Impact

• Development of suite of computer algorithms that enable computer simulations of the dynamics of epidemic growth at significantly cheaper computational costs, both in terms of computational time and computer memory requirements., Validate the developments through case studies involving existing epidemic growth models and available contact tracing networks.