Implements a filtering approach with a variational update based on Wasserstein gradient flows
Create a conda environment
conda create -n NAME python=3.9
Then head to the cloned repository and execute
pip install -e .
A filtering example on a stochastic volatility model
python examples/wasserstein_filter/markov_stochastic_volatility_wf_sqrt.py
@inproceedings{corenflos2023variational,
title={Variational Gaussian Filtering via Wasserstein Gradient Flows},
author={Corenflos, Adrien and Abdulsamad, Hany},
booktitle={2023 31st European Signal Processing Conference (EUSIPCO)},
year={2023},
}