GEM: a domain-specific modelling language for disease outbreak models
Project Overview
The GEM project is aiming to construct a domain-specific modelling language (DSML) for analysing infectious disease models for outbreaks such as Covid19, influenza, and malaria. The goal is to design a language that enables an epidemiologist to describe a disease
system, with the computer then automatically build the model, generating the statistical and machine learning algorithms for training (fitting) and prediction from that model. In doing so, we aim to revolutionise the speed at which epidemiologists can respond to the next disease outbreak, providing quantitative information for policy makers quickly and reliably.
What were the key results of your research project?
- Automated parsing and inference of simple Bayesian epidemic models
- Exploration of parameter correlation and its effects on mixing quality in MCMC algorithms
- Mitigation of parameter correlation through data manipulation and multi-site sampling
- Insight on the challenges of designing and implementing a DSML for infectious disease outbreaks
Github Repository: https://github.com/yycrystal/bayesian-epidemics-ir
A presentation of this research will be shared here and on our YouTube site when available.
How do you feel you have benefitted from completing this internship and has it made you consider future career paths?
This internship gave me a small preview of what research is like. Having agency in the direction of a project was a new experience for me, and I really valued it. It pushed me to think more independently. Working on a DSML for epidemic modelling showed me how applied mathematics translates into real-world tools. I started the project out of a passion for computational maths and to see whether research or further study could be the right path for me. I’m now confident it’s a path I want to pursue.
Download presentation slides