Estimating non-simultaneous changes in the mean of vectors (Q1669886)

In this paper the author presents change point estimators and their asymptotic distributions. A sequence of independent vectors with correlated components is considered. It is supposed that there is one change point in the mean of each component. Changes need not occur simultaneously. In an earlier paper [Test 24, No. 4, 681--700 (2015; Zbl 1329.62097)] the author has suggested a test statistic for detecting change points. Now in this paper change points are estimated by the argument at which this test statistic attains its maximum. These estimates are generalized least squares estimates. At first the author considers two-dimensional vectors and finds the asymptotic distribution of the change point estimate in the case of well separated change points (Theorem 1) and in the case of coinciding change points (Theorem 2). In Theorem 3 the well separated case is generalized for more than two dimensions.