Improving GW parameter-estimation using Gaussian process regression

Document #:

LIGO-G1501032-v2

Document type:

G - Presentations (eg Graphics)

Abstract:

Including uncertainty in theoretical models into Bayesian parameter estimation is necessary in order to make reliable inferences. A general means of achieving this is by marginalising over model uncertainty using a prior distribution established from Gaussian process regression (GPR). Here, we apply this technique to (simulated) gravitational-wave signals from binary black holes that could be observed using advanced-era gravitational-wave detectors. Uncertainty in the gravitational-wave templates could be the dominant source of error in studies of these systems unless they are correctly accounted for. We explain our approach in detail, providing proofs of various features
of the method, including the limiting behaviour for high signal-to-noise where systematic model uncertainties dominate over noise errors. We find that the marginalised likelihood offers a significant improvement in parameter estimation over the standard likelihood. We also examine the dependence of the method on the size of training set used in the GPR; the form of covariance function adopted for the GPR, and changes to the detector noise power spectral density.

Files in Document:

• Slides.pdf (file is not accessible)

Topics:

• Compact Binaries

Authors:

• Christopher Moore
• Christopher Berry
• Alvin Chua
• Jonathan Gair

Keywords:

Parameter-estimation Gaussian-process-regression

Related Documents:

• LIGO-P1500162: Improving gravitational-wave parameter estimation using Gaussian process regression

Publication Information:

http://arxiv.org/abs/1509.04066