Figure 3: False alarm rate versus detection statistic for compact binary coalescence (CBC) and burst
searches on 2009 – 2010 LIGO–Virgo data. Left: Cumulative rate of background events for a subset of
the CBC search parameter space, as a function of the threshold ranking statistic 
[19]. Different
methods were used to estimate the background for rate for high and low 
, which is why there is
an apparent gap in the data points. The background for the full search was approximately a factor
of six higher. Right: Cumulative rate of background events for the burst search, as a function of the
coherent network amplitude 
[14]. For ease of comparison, we have also plotted the approximate
equivalent 
for the burst search (an exact identification is not possible as the search methods
differ). The burst events are divided into two sets based on their central frequency.
[19]. Different
methods were used to estimate the background for rate for high and low , which is why there is
an apparent gap in the data points. The background for the full search was approximately a factor
of six higher. Right: Cumulative rate of background events for the burst search, as a function of the
coherent network amplitude [14]. For ease of comparison, we have also plotted the approximate
equivalent for the burst search (an exact identification is not possible as the search methods
differ). The burst events are divided into two sets based on their central frequency.