Quickest changepoint detection and identification for general stochastic models
Salle des séminaires (M.0.1)
LMRS, Univ. Rouen Normandie
In this talk we consider the problem of joint change detection and identification in the streams of the observations. We propose a changepoint detection-identification procedure that controls the probabilities of false alarm and wrong identification. We show that the proposed procedure is optimal in the minimax sense as probabilities of false alarm and wrong identification approach zero. The asymptotic optimality properties hold for general stochastic models with dependent and nonidentically distributed observations. We illustrate general results for detection-identification of changes in multistream Markov ergodic processes. We apply the constructed procedures to the rapid detection-identification problem of COVID-19 in Italy. Our proposed sequential algorithm allows much faster detection of COVID-19 than standard methods.