Epidemic intervention modelling laboratory
A mathematical modelling laboratory for studying how assumptions and selected interventions change simulated epidemic trajectories.
Project definition
Problem statement
Epidemic models simplify contact, transmission, recovery, reporting, immunity, and population structure. Different parameter combinations can fit the same observed curve but imply different futures.
The mathematical problem is to formulate model assumptions clearly, estimate which parameters are identifiable from available aggregate data, and propagate uncertainty through intervention comparisons.
Project objectives
- Implement selected SIR, SEIR, age-structured, or metapopulation variants under documented assumptions.
- Fit model parameters to synthetic or permitted aggregate time series with observation-error models.
- Analyse parameter identifiability, correlation, sensitivity, and uncertainty.
- Represent selected contact, detection, isolation, vaccination, or treatment scenarios mathematically.
- Compare trajectories and resource proxies without presenting them as real public-health forecasts.
System design
System modules
Model builder
Defines compartments, flows, population groups, initial conditions, parameters, and conservation equations.
Numerical solver
Integrates deterministic differential equations and selected stochastic simulations under recorded settings.
Parameter estimator
Fits bounded parameters to prepared observations and displays residuals, correlation, and uncertainty.
Intervention laboratory
Applies time-dependent parameter changes and compares selected timing, coverage, and effectiveness assumptions.
Analysis dashboard
Shows compartments, incidence, uncertainty bands, fits, sensitivities, interventions, and assumption tables.
Methodology
System workflow
- 01Choose model
The student selects a prepared population, compartment structure, observation process, and dataset.
- 02Verify equations
Mass conservation, non-negativity, units, and analytical special cases are checked.
- 03Estimate parameters
Selected parameters fit the training period with bounds, priors or penalties, and uncertainty.
- 04Run interventions
Prepared scenarios change documented model parameters over selected periods.
- 05Compare sensitivity
Trajectories are interpreted alongside parameter ranges, identifiability, and alternative assumptions.
Demonstration scenario
A synthetic outbreak is generated from a known SEIR model with incomplete reporting. The student fits several parameter sets that explain the observations similarly, then compares an earlier and later intervention. The dashboard shows how parameter uncertainty widens both scenario trajectories.
Engineering
Technical architecture
- Web application
- Nuxt and Vue for model diagrams, parameter controls, fit plots, scenario comparison, and exports.
- Modelling API
- FastAPI for equations, solver jobs, fitting, uncertainty, sensitivity, scenarios, and results.
- Data layer
- PostgreSQL for models, parameter sets, aggregate observations, fits, scenarios, and experiment records.
- Numerical engine
- Python, NumPy, and SciPy for differential equations, stochastic runs, constrained fitting, and sensitivity.
- Verification
- Conservation tests, known analytical cases, synthetic parameter recovery, solver tolerance, and repeated simulations.
Testing
Evaluation
Evaluation measures
- Equation conservation, non-negativity, and solver convergence
- Recovery of known parameters from synthetic observations
- Fit error and residual behaviour on prepared aggregate data
- Parameter confidence, correlation, practical identifiability, and sensitivity
- Uncertainty coverage for held-out synthetic trajectories
- Solver, fitting, and simulation time across structures and scenario counts
System boundaries
- The laboratory is for mathematical education and does not predict a current outbreak or recommend public-health action.
- Only synthetic, public aggregate, or explicitly permitted de-identified data is included.
- Model outputs depend strongly on structure, reporting, behaviour, intervention, and parameter assumptions.
- Real decisions require current surveillance, domain experts, ethical review, and responsible public-health authorities.
Included
- 01Population, model, parameter, and intervention interface
- 02Differential-equation, fitting, uncertainty, and sensitivity modules
- 03Compartment, incidence, parameter, scenario, and uncertainty views
- 04Prepared synthetic or aggregate public datasets and results
- 05Complete source code in a private GitHub repository
- 06Complete project documentation with synopsis, abstract, methodology, mathematical formulation, simulation results, screenshots, and conclusion
- 07Setup and usage guide
Project record
No buyer information is collected on this page.
- Permanent project ID
- GP-MA-0X9O2PO
- Catalogued
- 21 Aug 2026
- Completed
- Pending
- Verified
- Pending
- Demonstration
- Added when ready
Handover
After purchase
- 01Payment is confirmed
The project is marked unavailable and cannot be purchased again.
- 02Repository access is granted
The buyer's submitted GitHub account receives access to the private repository.
- 03The purchase record is delivered
The certification sheet is prepared from the reviewed buyer details and sent privately by email.