论文信息 - Robust Jump Detection in Regression Surface

Robust Jump Detection in Regression Surface

We use the difference of two asymmetric M-kernel estimators to detect jumps in two-dimensional regression functions. The method extends and corrects the Rotational Difference Kernel Estimator method proposed by Qiu (1997). For regression functions with only one explicit jump curve and additive noise, we show consistency for the jump location and height. In a simulation study, the consistency is also demonstrated for the case that 30% of the observations are replaced by outliers. In this case, the robust M-kernel estimators are superior to the classical kernel-estimators.

C. Müller

[1] Larry S. Davis,et al. A survey of edge detection techniques , 1975 .

[2] David Malah,et al. A study of edge detection algorithms , 1982, Comput. Graph. Image Process..

[3] W. Härdle,et al. Robust Non-parametric Function Fitting , 1984 .

[4] M. C. Jones,et al. Spline Smoothing and Nonparametric Regression. , 1989 .

[5] P. Qiu,et al. ESTIMATION OF JUMP REGRESSION FUNCTION , 1991 .

[6] H. Müller. CHANGE-POINTS IN NONPARAMETRIC REGRESSION ANALYSIS' , 1992 .

[7] C. Chu,et al. Kernel-Type Estimators of Jump Points and Values of a Regression Function , 1993 .

[8] A. Tsybakov,et al. Minimax theory of image reconstruction , 1993 .

[9] H. Müller,et al. Maximin estimation of multidimensional boundaries , 1994 .

[10] B. Yandell,et al. Jump Detection in Regression Surfaces , 1997 .

[11] P. Qiu. A Nonparametric Procedure to Detect Jumps in Regression Surfaces , 2002 .

[12] Zujun Hou,et al. Robust edge detection , 2003, Pattern Recognit..

[13] D. Ferger. Boundary estimation based on set-indexed empirical processes , 2004 .

[14] A. Cuevas,et al. On boundary estimation , 2004, Advances in Applied Probability.