Robust Bayesian detection of unmodelled bursts

TL;DR

This paper applies Bayesian methods to detect unmodelled gravitational wave bursts, revealing biases in existing statistics and proposing a more optimal detection approach for realistic signals.

Contribution

It introduces a Bayesian framework for unmodelled burst detection, highlighting biases in prior methods and improving detection performance.

Findings

Identifies biases in existing detection statistics.

Proposes a Bayesian detection method.

Demonstrates improved detection with realistic signals.

Abstract

A Bayesian treatment of the problem of detecting an unmodelled gravitational wave burst with a global network of gravitational wave observatories reveals that several previously proposed statistics have implicit biases that render them sub-optimal for realistic signal populations.