摘要

Along the lines of Huskova and Koubkova [Sequential procedures for detection of changes in autoregressive sequences, in Proceedings of Prague Stochastics 2006, M. Huskova and M. Janzura, eds., MatfyzPress, Charles University, Prague, pp. 437447], we further investigate a sequential procedure for monitoring jump changes in linear models. Our main result shows that, under the alternative, the suitably normalized stopping time of the procedure has a standard normal limiting distribution. A number of examples are discussed and the finite sample validity of the asymptotics is checked via a small simulation study.

  • 出版日期2013-2-1