Non-Linear Models for Extremal Dependence

The dependence structure of max-stable random vectors can be characterized by their Pickands dependence function. In many applications, the extremal dependence measure varies with covariates. We develop a flexible, semi-parametric method for the estimation of non-stationary multivariate Pickands dependence functions. The proposed construction is based on an accurate max-projection allowing to pass from the multivariate to the univariate setting and to rely on the generalized additive modeling framework. In the bivariate case, the resulting estimator of the Pickands function is regularized using constrained median smoothing B-splines, and bootstrap variability bands are constructed. In higher dimensions, we tailor our approach to the estimation of the extremal coefficient. An extended simulation study suggests that our estimator performs well and is competitive with the standard estimators in the absence of covariates. We apply the new methodology to a temperature dataset in the US where the extremal dependence is linked to time and altitude.

[1]  Christian Genest,et al.  Using B-splines for nonparametric inference on bivariate extreme-value copulas , 2014 .

[2]  S. Coles,et al.  An Introduction to Statistical Modeling of Extreme Values , 2001 .

[3]  Anthony C. Davison,et al.  Extremes on river networks , 2015, 1501.02663.

[4]  Paul Deheuvels,et al.  On the limiting behavior of the Pickands estimator for bivariate extreme-value distributions , 1991 .

[5]  L. de Haan,et al.  A Spectral Representation for Max-stable Processes , 1984 .

[6]  Jonathan A. Tawn,et al.  A dependence measure for multivariate and spatial extreme values: Properties and inference , 2003 .

[7]  Projection estimators of Pickands dependence functions , 2008 .

[8]  B. Renard,et al.  Use of a Gaussian copula for multivariate extreme value analysis: Some case studies in hydrology , 2007 .

[9]  S. Resnick.  Extreme Values, Regular Variation, and Point Processes , 1987 .

[10]  Anthony C. Davison,et al.  Statistics of Extremes , 2015 .

[11]  A. Guillou,et al.  Local robust estimation of the Pickands dependence function , 2018, The Annals of Statistics.

[12]  G. Marcon,et al.  Multivariate nonparametric estimation of the Pickands dependence function using Bernstein polynomials , 2014 .

[13]  Marc G. Genton,et al.  Non-Stationary Dependence Structures for Spatial Extremes , 2014, 1411.3174.

[14]  H. Drees,et al.  Best Attainable Rates of Convergence for Estimators of the Stable Tail Dependence Function , 1998 .

[15]  Martin Schlather,et al.  Models for Stationary Max-Stable Random Fields , 2002 .

[16]  Pin T. Ng,et al.  COBS: qualitatively constrained smoothing via linear programming , 1999, Comput. Stat..

[17]  Johan Segers,et al.  Nonparametric estimation of multivariate extreme-value copulas , 2011, 1107.2410.

[18]  J. Tawn.  Modelling multivariate extreme value distributions , 1990 .

[19]  S. Coles,et al.  Modelling Extreme Multivariate Events , 1991 .

[20]  Paul Embrechts,et al.  An Extreme Value Approach for Modeling Operational Risk Losses Depending on Covariates , 2016 .

[21]  Jonathan A. Tawn,et al.  Inequalities for the Extremal Coefficients of Multivariate Extreme Value Distributions , 2002 .

[22]  Paul Ressel.  Homogeneous distributions - And a spectral representation of classical mean values and stable tail dependence functions , 2013, J. Multivar. Anal..

[23]  Giampiero Marra,et al.  Practical variable selection for generalized additive models , 2011, Comput. Stat. Data Anal..

[24]  A. Davison,et al.  Generalized additive modelling of sample extremes , 2005 .

[25]  Pin T. Ng,et al.  A fast and efficient implementation of qualitatively constrained quantile smoothing splines , 2007 .

[26]  Michael G. Akritas,et al.  Recent Advances and Trends in Nonparametric Statistics , 2003 .

[27]  Philip Jonathan,et al.  Return level estimation from non-stationary spatial data exhibiting multidimensional covariate effects , 2014 .

[28]  Pin T. Ng,et al.  Quantile smoothing splines , 1994 .

[29]  Christian Genest,et al.  Multivariate Archimax copulas , 2014, J. Multivar. Anal..

[30]  Johan Segers,et al.  Nonparametric estimation of an extreme-value copula in arbitrary dimensions , 2009, J. Multivar. Anal..

[31]  S. Wood.  Generalized Additive Models: An Introduction with R , 2006 .

[32]  R. Tibshirani,et al.  Generalized additive models for medical research , 1986, Statistical methods in medical research.

[33]  P. Hall,et al.  Distribution and dependence-function estimation for bivariate extreme-value distributions , 2000 .

[34]  Anthony C. Davison,et al.  Modelling Time Series Extremes , 2012 .

[35]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[36]  J. Segers,et al.  RANK-BASED INFERENCE FOR BIVARIATE EXTREME-VALUE COPULAS , 2007, 0707.4098.

[37]  M. Parlange,et al.  Statistics of extremes in hydrology , 2002 .

[38]  Christian Genest,et al.  A nonparametric estimation procedure for bivariate extreme value copulas , 1997 .

[39]  Anthony C. Davison,et al.  Spectral Density Ratio Models for Multivariate Extremes , 2014 .

[40]  Brahim Brahimi.  Statistics of Bivariate Extreme Values , 2014 .