Skip to content

Latest commit

 

History

History
4 lines (3 loc) · 562 Bytes

README.md

File metadata and controls

4 lines (3 loc) · 562 Bytes

GaussianParticleFlow

An implementaion of the Gaussian flow method of importance sampling via optimal transport/SDE approximation. Used only for conditional non-linear Gaussian systems with known observational variance and non-linear transformation function. The original paper is: Pete Bunch, Simon Godsill, "Approximations of the Optimal Importance Density Using Gaussian Particle Flow Importance Sampling", Journal of American Statistical Association 2016, https://doi.org/10.1080/01621459.2015.1038387

Future work: clean up and documentation is required.