Joint segmentation of multivariate Gaussian processes using mixed linear models
Résumé
We consider the joint segmentation of multiple series. We use a mixed linear model to account for both covariates and correlations between signals. We propose an estimation algorithm based on EM which involves dynamic programming for the segmentation step. We show the computational e±ciency of this procedure. An application to microarray CGH profiles from multiple patients is presented.
Origine : Fichiers éditeurs autorisés sur une archive ouverte
Loading...