MG14 - Talk detail |
Participant |
Vetrugno, Daniele | |||||||
Institution |
University of Trento - via Sommarive, 14 - Trento - - Italy | |||||||
Session |
GW3 |
Accepted |
|
Order |
Time |
|||
Talk |
Oral abstract |
Title |
Fitting in unknown noise and applications to LISA Pathfinder | |||||
Coauthors | ||||||||
Abstract |
We proposed a new fitting technique when the background noise is unknown or a model for it is not accessible [Vitale et al. Physical Review D 90, 042003 (2014)]. The method is straightforward and is based on the analytical marginalization of the posterior probability. It returns the estimation of the best fit parameters as well as a consistent and unbiased measure of the background noise Power Spectral Density (PSD). This feature is particularly relevant for LISA Pathfinder for which noise PSD estimation is the main mission target. Moreover, we proved that this technique is equivalent to some implementations of the iterative reweighted least squares fit method, which is a very convenient algorithm when fitting linear models. Here, we report the results of applying the method to the LISA Pathfinder data and discuss its application to the gravitational wave detectors data analysis. |
|||||||
Pdf file |
||||||||