Get our free extension to see links to code for papers anywhere online!

Chrome logo  Add to Chrome

Firefox logo Add to Firefox


Ensemble transport smoothing -- Part 1: unified framework

Add code

Oct 31, 2022
Maximilian Ramgraber, Ricardo Baptista, Dennis McLaughlin, Youssef Marzouk


Share this with someone who'll enjoy it:


Smoothers are algorithms for Bayesian time series re-analysis. Most operational smoothers rely either on affine Kalman-type transformations or on sequential importance sampling. These strategies occupy opposite ends of a spectrum that trades computational efficiency and scalability for statistical generality and consistency: non-Gaussianity renders affine Kalman updates inconsistent with the true Bayesian solution, while the ensemble size required for successful importance sampling can be prohibitive. This paper revisits the smoothing problem from the perspective of measure transport, which offers the prospect of consistent prior-to-posterior transformations for Bayesian inference. We leverage this capacity by proposing a general ensemble framework for transport-based smoothing. Within this framework, we derive a comprehensive set of smoothing recursions based on nonlinear transport maps and detail how they exploit the structure of state-space models in fully non-Gaussian settings. We also describe how many standard Kalman-type smoothing algorithms emerge as special cases of our framework. A companion paper explores the implementation of nonlinear ensemble transport smoothers in greater depth.

* 31 pages, 13 figures, 4 pseudo-codes 


   Access Paper Source



Share this with someone who'll enjoy it: