ON GOODNESS-OF-FIT TESTING FOR CONTINUOUS TIME PROCESSES

Yury A. KUTOYANTS

Université du Maine, Le Mans
France


ABSTRACT

We present a review of some goodness-of-fit tests for diffusion and Poisson processes. Diffusion models are studied in small noise and large samples asymptotics. The basic hypothesis, as in classical statistics, is always simple and the tests are distribution-free. The power functions of the tests are described in the case of close nonparametric alternatives and the asymptotic optimality of some tests is discussed.



Back to the Workshop's Program