CVaR distance between univariate probability distributions and approximation problems

The paper defines new distances between univariate probability distributions, based on the concept of the CVaR norm. We consider the problem of approximation of a discrete distribution by some other discrete distribution. The approximating distribution has a smaller number of atoms than the original one. Such problems, for instance, must be solved for generation of scenarios in stochastic programming. The quality of the approximation is evaluated with new distances suggested in this paper. We use CVaR constraints to assure that the approximating distribution has tail characteristics similar to the target distribution. The numerical algorithm is based on two main steps: (i) optimal placement of positions of atoms of the approximating distribution with fixed probabilities; (ii) optimization of probabilities with fixed positions of atoms. These two steps are iterated to find both optimal atom positions and probabilities. Numerical experiments show high efficiency of the proposed algorithms, solved with convex and linear programming.

[1]  D. Darling.  The Kolmogorov-Smirnov, Cramer-von Mises Tests , 1957 .

[2]  R. Rockafellar,et al.  Optimization of conditional value-at risk , 2000 .

[3]  Dennis D. Boos,et al.  Minimum Distance Estimators for Location and Goodness of Fit , 1981 .

[4]  W. Rudin.  Principles of mathematical analysis , 1964 .

[5]  D. Keefer,et al.  Three-Point Approximations for Continuous Random Variables , 1983 .

[6]  S. Rachev,et al.  Advanced Stochastic Models, Risk Assessment, and Portfolio Optimization: The Ideal Risk, Uncertainty, and Performance Measures , 2008 .

[7]  Stanislav Uryasev,et al.  Conditional Value-at-Risk for General Loss Distributions , 2002 .

[8]  Johannes O. Royset,et al.  Random variables, monotone relations, and convex analysis , 2014, Math. Program..

[9]  Mircea Grigoriu,et al.  Reduced order models for random functions. Application to stochastic problems , 2009 .

[10]  James E. Smith.  Moment Methods for Decision Analysis , 1993 .

[11]  J. H. Schuenemeyer,et al.  A Modified Kolmogorov-Smirnov Test Sensitive to Tail Alternatives , 1983 .

[12]  John N. McDonald,et al.  A course in real analysis , 1999 .

[13]  Stan Uryasev,et al.  CVaR (superquantile) norm: Stochastic case , 2016, Eur. J. Oper. Res..

[14]  A. C. Miller,et al.  Discrete Approximations of Probability Distributions , 1983 .

[15]  Philippe Artzner,et al.  Coherent Measures of Risk , 1999 .

[16]  R. Rockafellar,et al.  Conditional Value-at-Risk for General Loss Distributions , 2001 .

[17]  G. Pflug.  Some Remarks on the Value-at-Risk and the Conditional Value-at-Risk , 2000 .

[18]  R. Rockafellar,et al.  The fundamental risk quadrangle in risk management, optimization and statistical estimation , 2013 .

[19]  Stan Uryasev,et al.  Probabilistic Constrained Optimization , 2000 .

[20]  J. Eric Bickel,et al.  Reexamining Discrete Approximations to Continuous Distributions , 2013, Decis. Anal..

[21]  Stan Uryasev,et al.  CVaR norm and applications in optimization , 2014, Optim. Lett..

[22]  C. Villani.  Optimal Transport: Old and New , 2008 .

[23]  Subhabrata Chakraborti,et al.  Nonparametric Statistical Inference , 2011, International Encyclopedia of Statistical Science.

[24]  R. Rockafellar.  Convex Analysis: (pms-28) , 1970 .

[25]  D. Keefer.  Certainty Equivalents for Three-Point Discrete-Distribution Approximations , 1994 .

[26]  S. S. Vallender.  Calculation of the Wasserstein Distance Between Probability Distributions on the Line , 1974 .

[27]  W. Feller.  On the Kolmogorov–Smirnov Limit Theorems for Empirical Distributions , 1948 .

[28]  Włodzimierz Ogryczak,et al.  On Robust Solutions to Multi-Objective Linear Programs , 2010 .

[29]  Karmeshu,et al.  Maximum entropy and discretization of probability distributions , 1987 .