HIGHLIGHTS
- What: The authors propose a novel localization strategy for the procedure based on sparse precision matrices which the authors denote precision localization. The authors evaluate the results by their prediction accuracy and to what degree the generated ensembles give a realistic representation of the exact filtering distributions. The authors propose a method that enables the authors to sample a sparse precision matrix, where the sparsity is predetermined by a graph. The authors propose to use it within an EnKF setting, whereas Carvalho and West discuss it within the framework of a dynamic linear model.
- Who . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.