Project Abstract

The primary motivation behind this research project is to enhance decision-making during public health emergencies, focusing on pandemic scenarios. This project addresses the limitations of such approaches by employing multi-objective Bayesian optimization, aiming to find a balance between disease control and economic impact. This is particularly crucial now, as recent experiences with COVID-19 have revealed gaps in real-time decision-making that considers long-term effects on public health and economies.This research is uniquely positioned to integrate advanced modeling frameworks that combine Bayesian methods with classical epidemiological models (e.g., the SIR model). The main aim is to develop a modeling framework that improves both the predictive accuracy of epidemiological forecasts and the economic impact analysis during pandemics. Bayesian Surrogate Models: These models are integrated within the classical SIR framework to handle uncertainties in disease transmission and recovery rates.Multi-Objective Bayesian Optimization (MOBO): This method is used to navigate the trade-offs between health outcomes (like infection rates) and economic impacts (such as GDP growth), aiming for optimal policy interventions.The developed models predict the spread of the virus and economic impacts with greater accuracy, providing a dual perspective on the outcomes of different policy decisions.By evaluating the trade-offs between public health and economic factors, the project provides essential insights that support nuanced policy decisions during pandemics.The research contributes significantly to our understanding of pandemic management by linking epidemiological models with economic impact assessments through a Bayesian probabilistic framework. It equips policymakers with a tool that enhances decision-making capacity in real-time, potentially saving lives and mitigating economic losses.

Keywords: Optimizing Public Health and Economic Outcomes, Bayesian Surrogate Models, Multi-Objective Bayesian Optimization (MOBO)

 Conference Details

Session: Presentation Stream 26 at Presentation Slot 8

Location: GH043 at Wednesday 8th 13:30 – 17:00

Markers: George Brooks (GTA), Markus Roggenbach

Course: MSc Data Science, Masters PG

Future Plans: I have a job lined-up