This repository contains scripts and code developed during my summer internship at the Swiss Tropical Health Institute (STPH) for building an emulator aimed at optimizing the OpenMalaria epidemiological model.
The code is organized to facilitate its use and understanding. It includes scripts for:
- Data Preparation: Filtering and structuring input and output data.
- Model Training: Scripts to train univariate and multivariate models (HetGP).
- Evaluation: Calculating performance metrics (MAPE, RMSE, R²) for predictions.
- Cluster Execution: Automation scripts for running training tasks on a compute cluster.
For detailed explanations on methods and technical choices, please consult the project’s Wiki.
- Clone the repository:
git clone https://github.com/username/repository_name.git
- Install dependencies:
# R, hetgp, and mlegp packages
Run the scripts by following the instructions in the Wiki. Options for training and model configuration are documented within the files.
- Matteo van Ypersele
- Supervised by Monica Golumbeanu and Christian Sellinger, STPH