[Previous] | [Session 26] | [Next]
P. C. Gregory (University of British Columbia)
The Bayesian view of probability theory as extended logic enables us to address the question of how to optimally detect periodic phenomena in astronomical time series for given state of prior knowledge of the signal shape and noise statistics. Bayesian methods give much more reliable and informative results and where the information warrants can lead to period estimates that have orders-of-magnitude greater resolution than conventional methods. My talk will focus on the application of probability theory to the detection and characterization of periodic signals where we have very little prior information of the signal shape and period. I will illustrate this with examples from X-ray and radio astronomy.
If the author provided an email address or URL for general inquiries, it is a
s follows: