Multiple change-points estimation of moving-average processes under dependence assumptions

In this paper, some results of convergence for a least-square estimator in the problem of multiple change-points estimation are presented and the moving-average processes of ρ-mixing sequence in the mean shifts are discussed. When the number of change points is known, the consistency of change-points estimator is derived. When the number of changes is unknown, the consistency of the change-points number and the change-points estimator by penalized least-squares method are obtained. The results are also true for φ-mixing, α-mixing, associated and negative associated sequences under suitable conditions.