The random Tukey depth

The computation of the Tukey depth, also called halfspace depth, is very demanding, even in low dimensional spaces, because it requires that all possible one-dimensional projections be considered. A random depth which approximates the Tukey depth is proposed. It only takes into account a finite number of one-dimensional projections which are chosen at random. Thus, this random depth requires a reasonable computation time even in high dimensional spaces. Moreover, it is easily extended to cover the functional framework. Some simulations indicating how many projections should be considered depending on the kind of problem, sample size and dimension of the sample space among others are presented. It is noteworthy that the random depth, based on a very low number of projections, obtains results very similar to those obtained with the Tukey depth.

[1]  Juan Romo,et al.  Depth-based classification for functional data , 2005, Data Depth: Robust Multivariate Analysis, Computational Geometry and Applications.

[2]  Frédéric Ferraty,et al.  Curves discrimination: a nonparametric functional approach , 2003, Comput. Stat. Data Anal..

[3]  V. Yohai,et al.  Robust Statistics: Theory and Methods , 2006 .

[4]  Ricardo Fraiman,et al.  On depth measures and dual statistics. A methodology for dealing with general data , 2009, J. Multivar. Anal..

[5]  Gérard Biau,et al.  On the Kernel Rule for Function Classification , 2006 .

[6]  David J. Hand,et al.  Classifier Technology and the Illusion of Progress , 2006, math/0606441.

[7]  P. Mahalanobis On the generalized distance in statistics , 1936 .

[8]  Florentina Bunea,et al.  Functional classification in Hilbert spaces , 2005, IEEE Transactions on Information Theory.

[9]  C. D. Kemp,et al.  Statistical Inference Based on Ranks , 1986 .

[10]  J. A. Cuesta-Albertos,et al.  A Sharp Form of the Cramér–Wold Theorem , 2007 .

[11]  Y. Zuo Multidimensional trimming based on projection depth , 2006, math/0702655.

[12]  Y. Zuo Projection-based depth functions and associated medians , 2003 .

[13]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[14]  P. Vieu,et al.  Nonparametric Functional Data Analysis: Theory and Practice (Springer Series in Statistics) , 2006 .

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

[16]  R. Fraiman,et al.  Trimmed means for functional data , 2001 .

[17]  Mia Hubert,et al.  Data Depth: Robust Multivariate Analysis, Computational Geometry and Applications by Liu, R., Serfling, R., and Souvaine, D. L. , 2008 .

[18]  Ricardo Fraiman,et al.  On the use of the bootstrap for estimating functions with functional data , 2006, Comput. Stat. Data Anal..

[19]  Karl Mosler,et al.  Data analysis and classification with the zonoid depth , 2003, Data Depth: Robust Multivariate Analysis, Computational Geometry and Applications.

[20]  J. Alexander,et al.  Theory and Methods: Critical Essays in Human Geography , 2008 .

[21]  J. A. Cuesta-Albertos,et al.  Random projections and goodness-of-fit tests in infinite-dimensional spaces , 2006 .

[22]  Juan Antonio Cuesta-Albertos,et al.  The random projection method in goodness of fit for functional data , 2007, Comput. Stat. Data Anal..

[23]  M. Febrero,et al.  Outlier detection in functional data by depth measures, with application to identify abnormal NOx levels , 2008 .

[24]  B. Silverman,et al.  Functional Data Analysis , 1997 .

[25]  Ricardo Fraiman,et al.  Robust estimation and classification for functional data via projection-based depth notions , 2007, Comput. Stat..

[26]  R. Serfling,et al.  General notions of statistical depth function , 2000 .