Reviewed by Lihong Wang References
Leif (d-gtn-st) Kempe, Angela Liebscher, Axel (d-gtn-st) Wittich +5 more
Consistencies and rates of convergence of jump-penalized least squares estimators. (English summary) Ann. Statist. 37 (2009), no. 1, 157–183. This paper studies the convergence of the jump-penalized least square estimators of the regression functions. The convergence rates of the estimators in the L 2 metric and the Skorokhod metric for the regression functions, and the convergence in the Hausdorff metric for the set of jump functions are both investigated. A simple data-driven parameter selection strategy is also introduced. The results can be extended to higher dimensions.