I'd actually prepped for this general topic so it wasn't a total blindside.
Start by clarifying the problem scope and assumptions, then propose a simple compartmental model (e.g., SIR) and iteratively extend it to address sub-parts like network effects, interventions, and scalability. Discuss trade-offs between model complexity and computational feasibility, and outline how you would validate and test the simulation.
Pro tip: Demonstrate awareness of real-world constraints by discussing how you would handle large populations (e.g., using approximate methods or parallelization) and how you would validate the model against historical data or known epidemiological properties.
Ask questions to understand the population structure, disease parameters, and expected outputs. Define assumptions about homogeneity, mixing, and time steps.
Select an appropriate model (e.g., SIR, SEIR, network-based) based on requirements. Explain the compartments, transitions, and equations governing the spread.
Outline the data structures (e.g., arrays for states, adjacency list for networks) and algorithm (e.g., discrete-time steps, event-driven). Discuss initialization and parameter setting.
For each sub-part (e.g., interventions, age structure, stochasticity), describe modifications to the model and simulation. Discuss how to incorporate them without redesigning from scratch.
Discuss computational complexity, potential optimizations (e.g., sampling, parallelization), and how to validate results (e.g., consistency checks, comparison with analytical solutions).
AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.