SPEAKER: Changliang Zou (visiting Dr. Jeff Wu)

ABSTRACT:

In multiple change-point problems, different data segments follow different distributions where changes may be in the mean, scale or the entire distribution from one segment to another. Without the need to know the number of change-points in advance, in this talk I will introduce a nonparametric maximum likelihood approach to detecting multiple change-points. Our method does not impose any parametric assumption on the underlying distributions of the data
sequence, which is thus suitable for detection of any changes in the distributions. The number of change-points is determined by the Bayesian information criterion and the locations of the
change-points can be estimated by using the dynamic programming algorithm and further taking advantage of the intrinsic order structure of the likelihood function. Under some mild conditions, we show that the new method provides consistent estimation for both the
locations and magnitudes of the change-points with a rate, the square of log n. We also suggest a pre-screening procedure which is capable of excluding most of the irrelevant points. Simulation studies show that the proposed method has outstanding performance of
identifying multiple change-points in terms of estimation accuracy and computation time compared with existing methods. The new methodology is illustrated with two real data examples.

Contact: Changliang ZOU <chlzou@yahoo.com.cn>