Tracking Stopping Times
Urs Niesen, Aslan Tchamkerten, Gregory W. Wornell · 2007
A novel quickest detection model is proposed which can be seen as a generalization of the Bayesian changepoint setup. Suppose {(Xi,Yi)}i≥1 is a sequence of pairs of random variables, and that S is a stopping time with respect to {Xi}i≥1. The problem is to find a stopping time T with respect to {Yi}i≥1 that optimally tracks S, in the sense that T minimizes the expected reaction time E(T − S) +, while keeps the false-alarm probability P(T