Imagine a child with a simple flu. A few days later, she is in the hospital with pneumonia. The flu didn’t cause it directly — it weakened her body just enough to let a second germ, pneumococcus, walk in through the open door.
This double attack is called co-infection, and it is my research topic.
So I ask: what if we could predict how these two diseases spread together? And then: with limited vaccines and a limited budget, who should get vaccinated, and when, to save the most lives?
I answer these questions with mathematics. I build a computer model of a whole population — sick and healthy, young and old, vaccinated and not — and test different vaccination plans on it, without risking a single real person.
Right now, I’m finishing the analysis and writing it up as a research paper. The goal is a smarter way to protect people every flu season.
In plain terms: it’s a mathematical way of finding the best possible strategyfor steering a system over time, when you can’t just fix everything at once — you have a limited budget, a limited supply, and every decision today affects what happens tomorrow. Here, the “system” is an entire population living through flu season, and the “strategy” is a vaccination plan.
The whole point is that this can be tested and re-tested entirely inside a computer — thousands of different vaccination plans can be tried instantly, and the mathematics guarantees you land on the genuinely best one, not just a plan that seems reasonable.
Smaller projects and reports from coursework across three institutions and two countries — pure mathematics, before the shift toward epidemiology and applied control theory.
I’m always happy to discuss mathematical epidemiology, optimal control, or the courses I teach at LUMS.