Likelihood of gravitational waves

Most discussion of gravitational wave detectors has hitherto been directed to optimization of the signal-to-noise ratio, leaving as a separate matter the question of interpretation of the observed waveforms, and particularly of putative 'events' observed simultaneously or with delay in individual detection systems. Maximization of signal-to-noise ratio is in fact only a means to the end of optimizing the modification of the probabilities assigned to hypotheses about gravitational radiation as a result of making the observations. Likelihood analysis appears able to tackle this problem directly in a way which automatically specifies optimal filtering and data-processing requirements.