Anonymous code release for review
This repository contains the implementation of BAD-PODS (Bayesian Adaptive Design for Partially Observable Dynamical Systems).
The code provides:
- Estimators of the expected information gain (EIG) and its gradient.
- An online framework based on nested particle filters (NPFs).
- Example experiments: two-group SIR model and moving source location.
This project is licensed under the MIT License.