Gaussian Process Emulates Complex Infectious Disease Models
Langmüller et al. (published December 29, 2025) develop an individual-based model inspired by dengue and train Gaussian Process surrogates to emulate outbreak probability, maximum incidence, and epidemic duration across an eight-dimensional parameter space. They calibrate the GP models using more than 1,000 dengue epidemics from 12 years of Colombian data, identify average infectivity and human mobility as dominant drivers, and show the emulators enable rapid prediction and location-specific inference.
Key Points
- 1Trained three Gaussian Process surrogates to predict outbreak probability, maximum incidence, and duration
- 2Showed average infectivity and human mobility drive epidemic outcomes in the eight-dimensional parameter space
- 3Enabled rapid predictions and calibration to Colombia dengue data, aiding location-specific inference
Scoring Rationale
Practical, well-validated emulation enabling fast, usable inference, with limited novelty beyond methodological application to dengue simulations.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
