Data depth for simple orthogonal regression with application to crack orientation

This paper studies tangential and simplicial data depth for simple orthogonal regression. Given N points in the plane, simple orthogonal regression means that we wish to determine the line through the origin that has smallest distance to the points measured in the direction orthogonal to the line. For both depth notions, it is proved that two lines which are orthogonal to each other, i.e. two lines forming a cross, have the same depth. Depth-based orthogonal regression can thus merely fit crosses, not lines. We investigated the robustness properties of maximum depth estimators using the notion of exact fit. Another topic the paper covers is the testing of the hypothesis that the data points form a cross-like pattern. After a simple transformation, such a test can be based on the biggest data depth. The paper discusses an application of this test for the investigation of stress fractures in materials.

[1]  L. P. Pook,et al.  Linear Elastic Fracture Mechanics for Engineers: Theory and Applications , 2000 .

[2]  Stanislav Katina,et al.  Calculation of simplicial depth estimators for polynomial regression with applications , 2007, Comput. Stat. Data Anal..

[3]  Regina Y. Liu On a Notion of Data Depth Based on Random Simplices , 1990 .

[4]  K. Mardia Statistics of Directional Data , 1972 .

[5]  Michael Besel,et al.  Damage accumulation of graded steel , 2010 .

[6]  Angelika Brückner-Foit,et al.  On the determination of material parameters in crack initiation laws , 2008 .

[7]  P. Sprent,et al.  Statistical Analysis of Circular Data. , 1994 .

[8]  Ihara,et al.  A stochastic damage accumulation model for crack initiation in high‐cycle fatigue , 2000 .

[9]  Peter Cloetens,et al.  Study of the interaction of a short fatigue crack with grain boundaries in a cast Al alloy using X-ray microtomography , 2003 .

[10]  Christine H. Müller,et al.  Tests for multiple regression based on simplicial depth , 2010, J. Multivar. Anal..

[11]  Xuming He TAIL BEHAVIOR OF REGRESSION ESTIMATORS AND THEIR BREAKDOWN POINTS , 1990 .

[12]  Alan J. Lee,et al.  U-Statistics: Theory and Practice , 1990 .

[13]  Regina Y. Liu,et al.  Regression depth. Commentaries. Rejoinder , 1999 .

[14]  Amos Race,et al.  Application of circular statistics in the study of crack distribution around cemented femoral components. , 2003, Journal of biomechanics.

[15]  Angelika Brückner-Foit,et al.  A stochastic simulation model for microcrack initiation in a martensitic steel , 2003 .

[16]  C. Müller Depth estimators and tests based on the likelihood principle with application to regression , 2005 .

[17]  Z. Bai,et al.  Asymptotic distributions of the maximal depth estimators for regression and multivariate location , 1999 .

[18]  M. Besel,et al.  Surface damage evolution of engineering steel , 2008 .

[19]  R. Y. Liu,et al.  On a notion of simplicial depth. , 1988, Proceedings of the National Academy of Sciences of the United States of America.

[20]  Christine H. Müller,et al.  Depth notions for orthogonal regression , 2010, J. Multivar. Anal..

[21]  Francis Franklin,et al.  Image analysis to reveal crack development using a computer simulation of wear and rolling contact fatigue , 2003 .

[22]  J. Tukey Mathematics and the Picturing of Data , 1975 .

[23]  C. Müller,et al.  Breakdown points and variation exponents of robust $M$-estimators in linear models , 1999 .

[24]  M. Hubert,et al.  The Deepest Regression Method , 2002 .

[25]  Christine H. Müller,et al.  Distribution-free tests for polynomial regression based on simplicial depth , 2009, J. Multivar. Anal..

[26]  I. Mizera On depth and deep points: a calculus , 2002 .

[27]  E. Batschelet Circular statistics in biology , 1981 .

[28]  Yoshihiko Hamamoto,et al.  A Method for Crack Detection on a Concrete Structure , 2006, 18th International Conference on Pattern Recognition (ICPR'06).

[29]  S. P. Ellis,et al.  Leverage and Breakdown in L 1 Regression , 1992 .

[30]  Shivprakash Iyer,et al.  A robust approach for automatic detection and segmentation of cracks in underground pipeline images , 2005, Image Vis. Comput..