The skew-normal distribution and related multivariate families. Adelchi Azzalini. Università di Padova. NordStat 2004: Jyväskylä, 6 10 June 2004
|
|
- Russell Page
- 5 years ago
- Views:
Transcription
1 The skew-normal distribution and related multivariate families Adelchi Azzalini Università di Padova NordStat 2004: Jyväskylä, 6 10 June 2004 A. Azzalini, NordStat 2004 p.1/53
2 Overview Background & motivation Skew-normal distribution (scalar/multivariate case, extensions) More general families skew-elliptical families, and some special cases a general formulation semi-parametric families Connection with other areas selective sampling stochastic frontier models forms of robust likelihood applications in mathematical geology, finance, etc. Some open problems A. Azzalini, NordStat 2004 p.2/53
3 Some general remarks A. Azzalini, NordStat 2004 p.3/53
4 Motivation Context: parametric families of pdf s for continuous variates Want to build flexible parametric classes extend normal family, in scalar and multivariate case construct families of arbitrary flexibility bridge between parametric and non-parametric approach normal class is central to multivariate data analysis,... but who has ever seen a multivariate normal sample? [Barnett (1979)] A. Azzalini, NordStat 2004 p.4/53
5 An old history (bivariate) Pearson family marginal transformation approach, including: Johnson family transformations and Edgeworth expansions mixtures and compound dimensional case: some approaches are specific of copulae specification via conditional distributions others... [Reviews: Pretorious (1930), Johnson & Kotz ( ), Hutchinson & Lay (1990), Joe (1997), Patil et al. (1976), Kotz et al. (2000 ),... ] A. Azzalini, NordStat 2004 p.5/53
6 Desiderata high flexibility in shape with wide range for indices of skewness and kurtosis mathematically tractable: simple form of density, moments, cdf, etc. closed under marginalization and conditioning retain nice properties of original family parameters should be directly linked to some aspects of pdf association with physical mechanism(s) to support statistical modeling for random number generation allow easy and reliable inference A. Azzalini, NordStat 2004 p.6/53
7 Expanding the Normal family: an approach A. Azzalini, NordStat 2004 p.7/53
8 $) A simple yet useful result Lemma (1D case) symmetric about 0, then are pdf s on and Suppose. is a density for any odd function are independent, and Proof. If ] &(' #%$ " [Azzalini (SJS, 1985) for case A. Azzalini, NordStat 2004 p.8/53
9 ) The skew-normal distribution ( Use lemma to skew the N(0,1) density:. Write for Some basic facts:, obtain N(0,1) if, then if for its cdf Extended version: A. Azzalini, NordStat 2004 p.9/53
10 Some SN pdf s ( ) α = α = pdf SN(0,1, 5) pdf SN(0,1,5) α = 0 α = A. Azzalini, NordStat 2004 p.10/53 pdf SN(0,1,0) pdf SN(0,1,Inf)
11 Forms of genesis, correlation and marginals N(0,1) with Conditioning: if then For Sum of Normal+HN: if iid, then Order statistics: A. Azzalini, NordStat 2004 p.11/53
12 A constructive example group above conditional mean SN Normal W Density H W W>E{W H} separate out group with W above average (given H): A. Azzalini, NordStat 2004 p.12/53
13 Some nice properties is Owen s function,, independent of hence even moments are the same of HN representation odd moments from N, or use (scaled) etc... [Azzalini (1985), Henze (1986), Chiogna (1998)] A. Azzalini, NordStat 2004 p.13/53
14 Early (independent) appearances Conditioning: Birnbaum (1950) obtains pdf of and some properties of first two moments. Motivation: selection process of personnel, in education, etc. Sum of Norm+HN: in Technom. (1964), discussion with contributions by Weinstein, Lipow, Mantel, Wilkinson and Nelson in Bayesian context, O Hagan & Leonard (1976) obtain ESN from a two-stage prior construction Order statistics: Roberts (1966) obtains SN pdf, and its property A. Azzalini, NordStat 2004 p.14/53
15 Other ways of skewing the normal pdf, use previous lemma with many options... [Nadarajah & Kotz (2003)] constant connnection with order statistics of N(0,1) [Balakrishnan (in discussion on Arnold & Beaver, 2002)] or take entirely different routes: two-pieces normal, etc. A. Azzalini, NordStat 2004 p.15/53
16 Skewing non-normal distributions ) exponential power dist n (or GED or Normal for heavy (or light) tails SEP [Azzalini (1986), DiCiccio & Monti (2004)] skew-laplace [Balakrishnan and Ambalagaspitya (1994)] skew-cauchy [Arnold & Beaver (2000b), include multivariate version] skew-logistic [Wahed & Ali (2000)] the wrapped SN on the circle (related to above form but not quite the same) [Pewsey (2000b, 2003)] A. Azzalini, NordStat 2004 p.16/53
17 Statistical aspects A. Azzalini, NordStat 2004 p.17/53
18 An old example: Johannsen s beans data Breadth of Phaseolus Vulgaris Johannsen s data -value d. f. Distribution normal log normal skew normal Normal Log-normal Edgeworth density 1st order nd order Skew-normal Note: data re-grouped from 20 into 16 bins [Charlier (1931), Cramér (1946, p )] breadth (mm) n=12000, source: Charlier, 1931 A. Azzalini, NordStat 2004 p.18/53
19 Likelihood function is regular, but... : always a stationarity point of Fisher expected information matrix becomes singular at [connection with non-standard cases of Rotnitzky et al. (2000)] Both can be avoided by suitable reparametrization for small MLE can occur at some solutions proposed MLE can have more than one local maximum small here is larger than elsewere [Azzalini (1985), Liseo (1990), Azzalini & Capitanio (1999), Pewsey (2000a), Monti (2003), Sartori (2003+) ] A. Azzalini, NordStat 2004 p.19/53
20 An example of profile log-likelihood dataset: otis dataset: otis lambda gamma omega Profile relative 2(logLikelihood) sigma Profile relative 2(logLikelihood) use of direct vs centred parameterization A. Azzalini, NordStat 2004 p.20/53
21 Bayesian inference ESN arises from two-stage prior [O Hagan & Leonard (1976)] Bayesian approach to estimation of parameters can avoid. [Liseo (1990), see also Liseo & Loperfido (2002, to appear)] linear Bayesian estimation for model [Mukhopadhyay & Vidakovic (1995)] A. Azzalini, NordStat 2004 p.21/53
22 Multivariate SN distribution A. Azzalini, NordStat 2004 p.22/53
23 Multivariate versions of 1D genesis Conditioning: marginals function Sum of Norm+HN: function The two mechanisms generate the same class of pdf s [Azzalini & Dalla Valle (1996)] A. Azzalini, NordStat 2004 p.23/53
24 Multivariate SN distribution is if pdf at Standard SN: write., get. If Add location/scale parameters: A. Azzalini, NordStat 2004 p.24/53
25 Some z z 1 pdf s α 1 = 2 α 2 = 6 Ω 12 = α 1 = 2 α 2 = Ω 12 = 0.7 α 1 = 2 α 2 = 6 Ω 12 = z2 z z z 1 A. Azzalini, NordStat 2004 p.25/53 α 1 = 2 α 2 = 6 Ω 12 = z z2
26 Some formal properties I are variation-independent Parameters Formal properties depend on e. g. var moment generating function & cumulants: -th order cumulant: constant A. Azzalini, NordStat 2004 p.26/53
27 Some formal properties II Random numbers: any of variate require a, namely computed via cdf of cdf of the class is closed under affine transformations hence also closed under marginalisation A. Azzalini, NordStat 2004 p.27/53
28 A characterisation property A parallel of a well-known property of the Normal class.. and var ) with be a r. v. (in. If Let Put., then such that for any [Gupta & Huang (2002)] A. Azzalini, NordStat 2004 p.28/53
29 Quadratic forms leads to Healy-type diagnostic plot, then and if, such that more general: for an even function for all, [Azzalini & Capitanio (1999, 2003), Genton et al. (2001), Loperfido (2001), Genton & Loperfido (to appear)] A. Azzalini, NordStat 2004 p.29/53
30 A numerical example: AIS data ) Australian Institute of Sport data ( observed Mahalanobis distances chi.square quantiles BMI scatterplot and SN fit using MLE LBM Healy s QQ plot for SN fit A. Azzalini, NordStat 2004 p.30/53
31 ESN and graphical models with pdf Extended SN: const is closed under conditioning ESN class [Arnold & Beaver (2000a)] Graphical models: can build a conditional independence graph for (with some restrictions) relationship between the conditional independence graphs of and of log-likelihood allows a parameter based factorization [Capitanio, Azzalini & Stanghellini (SJS, 2003)] A. Azzalini, NordStat 2004 p.31/53
32 Stochastic processes spatial processes with SN components, application to rainfall data, Bayesian inference via MCMC [Kim & Mallick (2004)] model for non-gaussian noise in signals processing [Gualtierotti (2004)] A. Azzalini, NordStat 2004 p.32/53
33 Multiple constraints where Consider Various cases: terms [Azzalini (1985), Arnold & Beaver (2000a) ] product of term, involves a case [Sahu, Dey & Branco (2004)] Hierarchical SN [Liseo & Loperfido (2003)] Fundamental SN [Arellano-Valle & Genton (to appear)] General SN ' [mentioned in Gupta, González-Farías & Domínguez-Molina (2004), with focus on case ] ' A. Azzalini, NordStat 2004 p.33/53
34 Perturbation of symmetric multivariate distributions A. Azzalini, NordStat 2004 p.34/53
35 Extend basic result to : (central) symmetry about 0 in N.B. includes elliptically contoured densities are pdf s symmetric about 0 in Lemma. Suppose and and, respectively, then, i. e. is a density for any real-valued odd function for.. Proof. Essentially the same of case, and other directions. -dimensional Can be extended to [Azzalini & Capitanio (2003), Genton & Loperfido (to appear), see also Arellano-Valle, del Pino & San Martín (2002) for another formulation] A. Azzalini, NordStat 2004 p.35/53
36 A simple stochastic representation, independent. Then, Suppose if if has density,, with and Corollary. If then. for any even function A. Azzalini, NordStat 2004 p.36/53
37 Can obtain high flexibility an example Consider a symmetric Beta pdf, scaled in : perturbed by where are additional parameters. There is more than skewing in the skew-symmetric class A. Azzalini, NordStat 2004 p.37/53
38 Perturbed symmetric Beta pdf s (a,b,p,q) = ( 2, 3, ( 3, 3 ), ( 0, 0 )) (a,b,p,q) = ( 2, 3, ( 8, 8 ), ( 0, 0 )) (a,b,p,q) = ( 3, 1, ( 1, 3 ), ( 2, 1 )) pdf pdf y y y x x x (a,b,p,q) = ( 3, 1.5, ( 3, 1 ), ( 2.5, 1 )) (a,b,p,q) = ( 3, 2, ( 2, 3 ), ( 2, 4 )) (a,b,p,q) = ( 3, 3, ( 1, 1 ), ( 3, 3 )) pdf pdf y y y x x x A. Azzalini, NordStat 2004 p.38/53 pdf pdf
39 Skew-ellipical densities and Route A: use previous lemma, with some [Azzalini & Capitanio (1999)] Route B: apply a random scale factor to a SN variate [Branco & Dey (2001)] Good news: for appropriate choice of the ingredients, essentially the same thing. [Azzalini & Capitanio (2003)] A. Azzalini, NordStat 2004 p.39/53
40 Analogies of SN and S-elliptical families the two stochastic representations for SN variates still hold, replacing normal with elliptical variates, via conditioning: via addition:, also representation via order statistics for case the distribution of quadratic forms is the same for skew-elliptical and elliptical variates with common symmetric component [Azzalini & Capitanio (2003), Fang (2003)] A. Azzalini, NordStat 2004 p.40/53
41 A noteworthy case: skew-, then and If has pdf where Unlimited range of indices of skewness and kurtosis. Suitable for robust inference. for QQ- and PP-plot A. Azzalini, NordStat 2004 p.41/53
42 Example: fiber glass strength ) data on breaking strength of fiber glass ( PP plot for skew t distribution PP plot for normal distribution Glass fiber data: nonparametric, SN and St fit glass glass nonparametric estimate, SN fit, ST fit Probability density function glass PP-plot for Normal model and ST model A. Azzalini, NordStat 2004 p.42/53
43 Example: MM returns ) MM excess returns vs CRSP index ( Histogram of res Density residuals CRSP linear regression with Normal and ST fit Martin Marietta returns histogram of residuals and fitted ST dist n A. Azzalini, NordStat 2004 p.43/53
44 & # & # Example: MM returns dataset: m.marietta Profile deviance PP plot for skew t distribution PP plot for normal distribution log(df) alpha m.marietta m.marietta ) for relative L PP-plot for Normal model and ST model profile A. Azzalini, NordStat 2004 p.44/53
45 Building expansions of arbitrary pdf s SkSy class: can be factorised as An arbitrary pdf is odd. where is symmetric around 0, and The factorisation depends on the chosen. denote odd polynomial of degree. The class Let is dense in SkSy-class (under some weak conditions). [ Wang, Boyer & Genton (2004), Ma & Genton (SJS, 2004) ] A. Azzalini, NordStat 2004 p.45/53
46 Some applications A. Azzalini, NordStat 2004 p.46/53
47 Use as a mathematical tool not really an application, still useful to approximate a distribution (see above) local departures from normality, local efficiency,... [Salvan (1986), Durio & Nikitin (2003), Muliere & Nikitin (2004)] to produce alternatives in simulation studies [DiCiccio et al. (1997), Chu et al. (2001), etc.] skew link for dicotomous response analysis [Chen, Dey & Qi (1999)] A. Azzalini, NordStat 2004 p.47/53
48 Connected areas Selected (non-random) sampling and Heckman model response eqn: selection eqn: corr type ESN [Copas & Li (1997)] stochastic frontier model [Aigner, Lowell & Schmidt (1977),..., Tancredi (2003+)] A. Azzalini, NordStat 2004 p.48/53
49 Compositional data i.e. frequent sort of data in geology, among others areas, popular approach: Aitchison ALR transformation to enjoyes a set of convenient formal properties s then apply methods for multivariate Normal data to preserves closure of suitable : and use of in place of transformations of on subcompositions [Mateu-Figueras (Ph.D thesis, 2003), Aitchison, Mateu-Figueras & Ng (2003)] A. Azzalini, NordStat 2004 p.49/53
50 Applications in finance multivariate SN fits well into portfolio selection theory use of skew- allows fat tails [Adcock & Shutes (1999), Adcock & Mead (2003), Harvey et al. (to appear)] incorporating skewness/kurtosis in components of ARCH, GARCH and SV models [Goria (thesis, 1999), Pietrobon (thesis, 2003), De Luca & Loperfido (2004), Cappuccio et al. (2004), Liseo & Loperfido (to appear)] A. Azzalini, NordStat 2004 p.50/53
51 nearly there... A. Azzalini, NordStat 2004 p.51/53
52 Some open problems a multitude of proposals (emphasis on probability side, less about statistics; what about their properties, connections, suitability for data analysis, etc.?) various issues in inference: avoid some peculiarities of MLE alternatives to MLE extend various methods from Normal to non-normal case can approximate any distribution, but what is best type of approximation? how to get approximation with prescribed properties? is all of this really useful? A. Azzalini, NordStat 2004 p.52/53
53 Resources References, some papers, software, etc. at A. Azzalini, NordStat 2004 p.53/53
Distributions generated by perturbation of symmetry with emphasis on a multivariate skew distribution
Distributions generated by perturbation of symmetry with emphasis on a multivariate skew distribution ADELCHI AZZALINI Dipartimento di Scienze Statistiche Università di Padova azzalini@stat.unipd.it ANTONELLA
More informationINFERENCE FOR MULTIPLE LINEAR REGRESSION MODEL WITH EXTENDED SKEW NORMAL ERRORS
Pak. J. Statist. 2016 Vol. 32(2), 81-96 INFERENCE FOR MULTIPLE LINEAR REGRESSION MODEL WITH EXTENDED SKEW NORMAL ERRORS A.A. Alhamide 1, K. Ibrahim 1 M.T. Alodat 2 1 Statistics Program, School of Mathematical
More informationOpening Skew-symmetric distn s Skew-elliptical distn s Other properties Closing. Adelchi Azzalini. Università di Padova, Italia
Stochastic representations and other properties of skew-symmetric distributions Adelchi Azzalini Università di Padova, Italia a talk at Universität Rostock, 29 Sept 2011 Overview a brief introduction to
More informationYUN WU. B.S., Gui Zhou University of Finance and Economics, 1996 A REPORT. Submitted in partial fulfillment of the requirements for the degree
A SIMULATION STUDY OF THE ROBUSTNESS OF HOTELLING S T TEST FOR THE MEAN OF A MULTIVARIATE DISTRIBUTION WHEN SAMPLING FROM A MULTIVARIATE SKEW-NORMAL DISTRIBUTION by YUN WU B.S., Gui Zhou University of
More informationarxiv: v1 [math.pr] 10 Oct 2017
Yet another skew-elliptical family but of a different kind: return to Lemma 1 arxiv:1710.03494v1 [math.p] 10 Oct 017 Adelchi Azzalini Dipartimento di Scienze Statistiche Università di Padova Italia Giuliana
More informationOn Bivariate Transformation of Scale Distributions
On Bivariate Transformation of Scale Distributions M.C. Jones The Open University, U.K. ABSTRACT Elsewhere, I have promoted the notion of (univariate continuous) transformation of scale (ToS) distributions
More informationSome properties of skew-symmetric distributions
Noname manuscript No. (will be inserted by the editor) Some properties of skew-symmetric distributions Adelchi Azzalini Giuliana Regoli 21st December 2010 Revised May 2011 Abstract The family of skew-symmetric
More informationJMASM25: Computing Percentiles of Skew- Normal Distributions
Journal of Modern Applied Statistical Methods Volume 5 Issue Article 3 --005 JMASM5: Computing Percentiles of Skew- Normal Distributions Sikha Bagui The University of West Florida, Pensacola, bagui@uwf.edu
More informationA study of generalized skew-normal distribution
Statistics, Vol. 00, No. 0, Month 2012, 1 12 0.8CE: VLD/SP QA: Coll: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
More informationEmpirical Bayes Estimation in Multiple Linear Regression with Multivariate Skew-Normal Distribution as Prior
Journal of Mathematical Extension Vol. 5, No. 2 (2, (2011, 37-50 Empirical Bayes Estimation in Multiple Linear Regression with Multivariate Skew-Normal Distribution as Prior M. Khounsiavash Islamic Azad
More informationSKEW-NORMALITY IN STOCHASTIC FRONTIER ANALYSIS
SKEW-NORMALITY IN STOCHASTIC FRONTIER ANALYSIS J. Armando Domínguez-Molina, Graciela González-Farías and Rogelio Ramos-Quiroga Comunicación Técnica No I-03-18/06-10-003 PE/CIMAT 1 Skew-Normality in Stochastic
More informationA NEW CLASS OF SKEW-NORMAL DISTRIBUTIONS
A NEW CLASS OF SKEW-NORMAL DISTRIBUTIONS Reinaldo B. Arellano-Valle Héctor W. Gómez Fernando A. Quintana November, 2003 Abstract We introduce a new family of asymmetric normal distributions that contains
More informationDefault Bayesian Analysis of the Skew-Normal Distribution
Default Bayesian Analysis of the Skew-Normal Distribution Brunero Liseo Università di Roma La Sapienza Nicola Loperfido Università di Urbino August 23, 2002 Corresponding author: Dipartimento di studi
More informationThe Mixture Approach for Simulating New Families of Bivariate Distributions with Specified Correlations
The Mixture Approach for Simulating New Families of Bivariate Distributions with Specified Correlations John R. Michael, Significance, Inc. and William R. Schucany, Southern Methodist University The mixture
More informationTwo Weighted Distributions Generated by Exponential Distribution
Journal of Mathematical Extension Vol. 9, No. 1, (2015), 1-12 ISSN: 1735-8299 URL: http://www.ijmex.com Two Weighted Distributions Generated by Exponential Distribution A. Mahdavi Vali-e-Asr University
More informationFinancial Econometrics and Quantitative Risk Managenent Return Properties
Financial Econometrics and Quantitative Risk Managenent Return Properties Eric Zivot Updated: April 1, 2013 Lecture Outline Course introduction Return definitions Empirical properties of returns Reading
More informationTheory and Methods of Statistical Inference
PhD School in Statistics cycle XXIX, 2014 Theory and Methods of Statistical Inference Instructors: B. Liseo, L. Pace, A. Salvan (course coordinator), N. Sartori, A. Tancredi, L. Ventura Syllabus Some prerequisites:
More informationSample mean, covariance and T 2 statistic of the skew elliptical model
Journal of Multivariate Analysis 97 (2006) 1675 1690 www.elsevier.com/locate/jmva Sample mean, covariance and T 2 statistic of the skew elliptical model B.Q. Fang,1 Institute of Applied Mathematics, Academy
More informationTheory and Methods of Statistical Inference. PART I Frequentist theory and methods
PhD School in Statistics cycle XXVI, 2011 Theory and Methods of Statistical Inference PART I Frequentist theory and methods (A. Salvan, N. Sartori, L. Pace) Syllabus Some prerequisites: Empirical distribution
More informationIdentifiability problems in some non-gaussian spatial random fields
Chilean Journal of Statistics Vol. 3, No. 2, September 2012, 171 179 Probabilistic and Inferential Aspects of Skew-Symmetric Models Special Issue: IV International Workshop in honour of Adelchi Azzalini
More informationTheory and Methods of Statistical Inference. PART I Frequentist likelihood methods
PhD School in Statistics XXV cycle, 2010 Theory and Methods of Statistical Inference PART I Frequentist likelihood methods (A. Salvan, N. Sartori, L. Pace) Syllabus Some prerequisites: Empirical distribution
More informationA BIMODAL EXPONENTIAL POWER DISTRIBUTION
Pak. J. Statist. Vol. 6(), 379-396 A BIMODAL EXPONENTIAL POWER DISTRIBUTION Mohamed Y. Hassan and Rafiq H. Hijazi Department of Statistics United Arab Emirates University P.O. Box 7555, Al-Ain, U.A.E.
More informationA Skewed Look at Bivariate and Multivariate Order Statistics
A Skewed Look at Bivariate and Multivariate Order Statistics Prof. N. Balakrishnan Dept. of Mathematics & Statistics McMaster University, Canada bala@mcmaster.ca p. 1/4 Presented with great pleasure as
More informationSkewed Reflected Distributions Generated by the Laplace Kernel
AUSTRIAN JOURNAL OF STATISTICS Volume 38 (009), Number 1, 45 58 Skewed Reflected Distributions Generated by the Laplace Kernel M. Masoom Ali 1, Manisha Pal and Jungsoo Woo 3 1 Dept. of Mathematical Sciences,
More informationContinuous Univariate Distributions
Continuous Univariate Distributions Volume 1 Second Edition NORMAN L. JOHNSON University of North Carolina Chapel Hill, North Carolina SAMUEL KOTZ University of Maryland College Park, Maryland N. BALAKRISHNAN
More informationThe Kumaraswamy Skew-t Distribution and Its Related Properties
Bowling Green State University ScholarWorks@BGSU Mathematics and Statistics Faculty Publications Mathematics and Statistics 6-29-2017 The Kumaraswamy Skew-t Distribution and Its Related Properties Wei
More informationHANDBOOK OF APPLICABLE MATHEMATICS
HANDBOOK OF APPLICABLE MATHEMATICS Chief Editor: Walter Ledermann Volume VI: Statistics PART A Edited by Emlyn Lloyd University of Lancaster A Wiley-Interscience Publication JOHN WILEY & SONS Chichester
More informationTruncated-Exponential Skew-Symmetric Distributions
Truncated-Exponential Skew-Symmetric Distributions Saralees Nadarajah, Vahid Nassiri & Adel Mohammadpour First version: 15 December 2009 Research Report No. 19, 2009, Probability and Statistics Group School
More informationBias-corrected Estimators of Scalar Skew Normal
Bias-corrected Estimators of Scalar Skew Normal Guoyi Zhang and Rong Liu Abstract One problem of skew normal model is the difficulty in estimating the shape parameter, for which the maximum likelihood
More informationStat 5101 Lecture Notes
Stat 5101 Lecture Notes Charles J. Geyer Copyright 1998, 1999, 2000, 2001 by Charles J. Geyer May 7, 2001 ii Stat 5101 (Geyer) Course Notes Contents 1 Random Variables and Change of Variables 1 1.1 Random
More informationMultivariate Distributions
IEOR E4602: Quantitative Risk Management Spring 2016 c 2016 by Martin Haugh Multivariate Distributions We will study multivariate distributions in these notes, focusing 1 in particular on multivariate
More informationOn Fisher Information Matrices and Profile Log-Likelihood Functions in Generalized Skew-Elliptical Models
On Fisher Information Matrices and Profile Log-Likelihood Functions in Generalized Skew-Elliptical Models Christophe Ley and Davy Paindaveine E.C.A.R.E.S., Institut de Recherche en Statistique, and Département
More informationp-birnbaum SAUNDERS DISTRIBUTION: APPLICATIONS TO RELIABILITY AND ELECTRONIC BANKING HABITS
p-birnbaum SAUNDERS DISTRIBUTION: APPLICATIONS TO RELIABILITY AND ELECTRONIC BANKING 1 V.M.Chacko, Mariya Jeeja P V and 3 Deepa Paul 1, Department of Statistics St.Thomas College, Thrissur Kerala-681 e-mail:chackovm@gmail.com
More informationMULTIVARIATE SKEW-T DISTRIBUTIONS IN ECONOMETRICS AND ENVIRONMETRICS. A Dissertation YULIA V. MARCHENKO
MULTIVARIATE SKEW-T DISTRIBUTIONS IN ECONOMETRICS AND ENVIRONMETRICS A Dissertation by YULIA V. MARCHENKO Submitted to the Office of Graduate Studies of Texas A&M University in partial fulfillment of the
More informationBayesian Regression Linear and Logistic Regression
When we want more than point estimates Bayesian Regression Linear and Logistic Regression Nicole Beckage Ordinary Least Squares Regression and Lasso Regression return only point estimates But what if we
More informationThe Metalog Distributions
Keelin Reeds Partners 770 Menlo Ave., Ste 230 Menlo Park, CA 94025 650.465.4800 phone www.keelinreeds.com The Metalog Distributions A Soon-To-Be-Published Paper in Decision Analysis, and Supporting Website
More informationA new construction for skew multivariate distributions
Journal of Multivariate Analysis 95 005 33 344 www.elsevier.com/locate/jmva A new construction for skew multivariate distributions DipakK. Dey, Junfeng Liu Department of Statistics, University of Connecticut,
More informationSTA 4273H: Statistical Machine Learning
STA 4273H: Statistical Machine Learning Russ Salakhutdinov Department of Statistics! rsalakhu@utstat.toronto.edu! http://www.utstat.utoronto.ca/~rsalakhu/ Sidney Smith Hall, Room 6002 Lecture 7 Approximate
More informationOn Five Parameter Beta Lomax Distribution
ISSN 1684-840 Journal of Statistics Volume 0, 01. pp. 10-118 On Five Parameter Beta Lomax Distribution Muhammad Rajab 1, Muhammad Aleem, Tahir Nawaz and Muhammad Daniyal 4 Abstract Lomax (1954) developed
More informationOn Perturbed Symmetric Distributions Associated with the Truncated Bivariate Elliptical Models
Communications of the Korean Statistical Society Vol. 15, No. 4, 8, pp. 483 496 On Perturbed Symmetric Distributions Associated with the Truncated Bivariate Elliptical Models Hea-Jung Kim 1) Abstract This
More informationPATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 2: PROBABILITY DISTRIBUTIONS
PATTERN RECOGNITION AND MACHINE LEARNING CHAPTER 2: PROBABILITY DISTRIBUTIONS Parametric Distributions Basic building blocks: Need to determine given Representation: or? Recall Curve Fitting Binary Variables
More informationOn Asymptotic Distribution of Sample Central. Moments in Normal-Uniform Distribution
International Mathematical Forum, Vol. 8, 2013, no. 8, 395-399 On Asymptotic Distribution of Sample Central Moments in Normal-Uniform Distribution Narges Abbasi Department of Statistics, Payame Noor University,
More informationTail dependence for two skew slash distributions
1 Tail dependence for two skew slash distributions Chengxiu Ling 1, Zuoxiang Peng 1 Faculty of Business and Economics, University of Lausanne, Extranef, UNIL-Dorigny, 115 Lausanne, Switzerland School of
More informationChapter 2 Normal Distribution
Chapter Normal Distribution The normal distribution is one of the most important continuous probability distributions, and is widely used in statistics and other fields of sciences. In this chapter, we
More informationContinuous Univariate Distributions
Continuous Univariate Distributions Volume 2 Second Edition NORMAN L. JOHNSON University of North Carolina Chapel Hill, North Carolina SAMUEL KOTZ University of Maryland College Park, Maryland N. BALAKRISHNAN
More informationQuadratic forms in skew normal variates
J. Math. Anal. Appl. 73 (00) 558 564 www.academicpress.com Quadratic forms in skew normal variates Arjun K. Gupta a,,1 and Wen-Jang Huang b a Department of Mathematics and Statistics, Bowling Green State
More informationDiscussion of The Skew-normal
Published by Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA Vol 32: 189 198, 2005 Discussion of The Skew-normal MARC G. GENTON Department
More informationHakone Seminar Recent Developments in Statistics
Hakone Seminar Recent Developments in Statistics November 12-14, 2015 Hotel Green Plaza Hakone: http://www.hgp.co.jp/language/english/sp/ Organizer: Masanobu TANIGUCHI (Research Institute for Science &
More informationSubject CS1 Actuarial Statistics 1 Core Principles
Institute of Actuaries of India Subject CS1 Actuarial Statistics 1 Core Principles For 2019 Examinations Aim The aim of the Actuarial Statistics 1 subject is to provide a grounding in mathematical and
More informationAn Extended Weighted Exponential Distribution
Journal of Modern Applied Statistical Methods Volume 16 Issue 1 Article 17 5-1-017 An Extended Weighted Exponential Distribution Abbas Mahdavi Department of Statistics, Faculty of Mathematical Sciences,
More informationMultivariate Birnbaum-Saunders Distribution Based on Multivariate Skew Normal Distribution
Multivariate Birnbaum-Saunders Distribution Based on Multivariate Skew Normal Distribution Ahad Jamalizadeh & Debasis Kundu Abstract Birnbaum-Saunders distribution has received some attention in the statistical
More informationModified Kolmogorov-Smirnov Test of Goodness of Fit. Catalonia-BarcelonaTECH, Spain
152/304 CoDaWork 2017 Abbadia San Salvatore (IT) Modified Kolmogorov-Smirnov Test of Goodness of Fit G.S. Monti 1, G. Mateu-Figueras 2, M. I. Ortego 3, V. Pawlowsky-Glahn 2 and J. J. Egozcue 3 1 Department
More informationBayesian closed skew Gaussian inversion of seismic AVO data into elastic material properties
Bayesian closed skew Gaussian inversion of seismic AVO data into elastic material properties Omid Karimi 1,2, Henning Omre 2 and Mohsen Mohammadzadeh 1 1 Tarbiat Modares University, Iran 2 Norwegian University
More informationMultivariate Non-Normally Distributed Random Variables
Multivariate Non-Normally Distributed Random Variables An Introduction to the Copula Approach Workgroup seminar on climate dynamics Meteorological Institute at the University of Bonn 18 January 2008, Bonn
More informationOn the Conditional Distribution of the Multivariate t Distribution
On the Conditional Distribution of the Multivariate t Distribution arxiv:604.0056v [math.st] 2 Apr 206 Peng Ding Abstract As alternatives to the normal distributions, t distributions are widely applied
More informationON EXTREME VALUE ANALYSIS OF A SPATIAL PROCESS
REVSTAT Statistical Journal Volume 6, Number 1, March 008, 71 81 ON EXTREME VALUE ANALYSIS OF A SPATIAL PROCESS Authors: Laurens de Haan Erasmus University Rotterdam and University Lisbon, The Netherlands
More informationMultivariate Statistics
Multivariate Statistics Chapter 2: Multivariate distributions and inference Pedro Galeano Departamento de Estadística Universidad Carlos III de Madrid pedro.galeano@uc3m.es Course 2016/2017 Master in Mathematical
More informationMultivariate Asset Return Prediction with Mixture Models
Multivariate Asset Return Prediction with Mixture Models Swiss Banking Institute, University of Zürich Introduction The leptokurtic nature of asset returns has spawned an enormous amount of research into
More informationParametric Inference Maximum Likelihood Inference Exponential Families Expectation Maximization (EM) Bayesian Inference Statistical Decison Theory
Statistical Inference Parametric Inference Maximum Likelihood Inference Exponential Families Expectation Maximization (EM) Bayesian Inference Statistical Decison Theory IP, José Bioucas Dias, IST, 2007
More informationBayesian inference for multivariate skew-normal and skew-t distributions
Bayesian inference for multivariate skew-normal and skew-t distributions Brunero Liseo Sapienza Università di Roma Banff, May 2013 Outline Joint research with Antonio Parisi (Roma Tor Vergata) 1. Inferential
More informationPRINCIPLES OF STATISTICAL INFERENCE
Advanced Series on Statistical Science & Applied Probability PRINCIPLES OF STATISTICAL INFERENCE from a Neo-Fisherian Perspective Luigi Pace Department of Statistics University ofudine, Italy Alessandra
More informationFinancial Econometrics and Volatility Models Copulas
Financial Econometrics and Volatility Models Copulas Eric Zivot Updated: May 10, 2010 Reading MFTS, chapter 19 FMUND, chapters 6 and 7 Introduction Capturing co-movement between financial asset returns
More informationThe Instability of Correlations: Measurement and the Implications for Market Risk
The Instability of Correlations: Measurement and the Implications for Market Risk Prof. Massimo Guidolin 20254 Advanced Quantitative Methods for Asset Pricing and Structuring Winter/Spring 2018 Threshold
More informationCopula Regression RAHUL A. PARSA DRAKE UNIVERSITY & STUART A. KLUGMAN SOCIETY OF ACTUARIES CASUALTY ACTUARIAL SOCIETY MAY 18,2011
Copula Regression RAHUL A. PARSA DRAKE UNIVERSITY & STUART A. KLUGMAN SOCIETY OF ACTUARIES CASUALTY ACTUARIAL SOCIETY MAY 18,2011 Outline Ordinary Least Squares (OLS) Regression Generalized Linear Models
More informationPROBABILITY DISTRIBUTIONS. J. Elder CSE 6390/PSYC 6225 Computational Modeling of Visual Perception
PROBABILITY DISTRIBUTIONS Credits 2 These slides were sourced and/or modified from: Christopher Bishop, Microsoft UK Parametric Distributions 3 Basic building blocks: Need to determine given Representation:
More informationSkew-symmetric distributions and Fisher information: The double sin of the skew-normal
Bernoulli 2(3), 214, 1432 1453 DOI: 1.315/13-BEJ528 arxiv:129.4177v3 [math.st] 9 Jul 214 Skew-symmetric distributions and Fisher information: The double sin of the skew-normal MARC HALLIN 1 and CHRISTOPHE
More informationBivariate Degradation Modeling Based on Gamma Process
Bivariate Degradation Modeling Based on Gamma Process Jinglun Zhou Zhengqiang Pan Member IAENG and Quan Sun Abstract Many highly reliable products have two or more performance characteristics (PCs). The
More informationSTATISTICS SYLLABUS UNIT I
STATISTICS SYLLABUS UNIT I (Probability Theory) Definition Classical and axiomatic approaches.laws of total and compound probability, conditional probability, Bayes Theorem. Random variable and its distribution
More informationMotivation Scale Mixutres of Normals Finite Gaussian Mixtures Skew-Normal Models. Mixture Models. Econ 690. Purdue University
Econ 690 Purdue University In virtually all of the previous lectures, our models have made use of normality assumptions. From a computational point of view, the reason for this assumption is clear: combined
More informationA weather generator for simulating multivariate climatic series
A weather generator for simulating multivariate climatic series Denis Allard, with Nadine Brisson (AgroClim, INRA), Cédric Flecher (MetNext) and Philippe Naveau (LSCE, CNRS) Biostatistics and Spatial Processes
More informationTail Approximation of the Skew-Normal by the Skew-Normal Laplace: Application to Owen s T Function and the Bivariate Normal Distribution
Journal of Statistical and Econometric ethods vol. no. 3 - ISS: 5-557 print version 5-565online Scienpress Ltd 3 Tail Approimation of the Skew-ormal by the Skew-ormal Laplace: Application to Owen s T Function
More informationFrom Gram-Charlier Series to Orthogonal Polynomials
From Gram-Charlier Series to Orthogonal Polynomials HC Eggers University of Stellenbosch Workshop on Particle Correlations and Fluctuations 2008 Kraków Earlier work on Gram-Charlier Series (GCS) Work by
More informationTail dependence in bivariate skew-normal and skew-t distributions
Tail dependence in bivariate skew-normal and skew-t distributions Paola Bortot Department of Statistical Sciences - University of Bologna paola.bortot@unibo.it Abstract: Quantifying dependence between
More informationFirst steps of multivariate data analysis
First steps of multivariate data analysis November 28, 2016 Let s Have Some Coffee We reproduce the coffee example from Carmona, page 60 ff. This vignette is the first excursion away from univariate data.
More informationSTATISTICS ANCILLARY SYLLABUS. (W.E.F. the session ) Semester Paper Code Marks Credits Topic
STATISTICS ANCILLARY SYLLABUS (W.E.F. the session 2014-15) Semester Paper Code Marks Credits Topic 1 ST21012T 70 4 Descriptive Statistics 1 & Probability Theory 1 ST21012P 30 1 Practical- Using Minitab
More informationCurve Fitting Re-visited, Bishop1.2.5
Curve Fitting Re-visited, Bishop1.2.5 Maximum Likelihood Bishop 1.2.5 Model Likelihood differentiation p(t x, w, β) = Maximum Likelihood N N ( t n y(x n, w), β 1). (1.61) n=1 As we did in the case of the
More informationModulation of symmetric densities
1 Modulation of symmetric densities 1.1 Motivation This book deals with a formulation for the construction of continuous probability distributions and connected statistical aspects. Before we begin, a
More informationECE521 week 3: 23/26 January 2017
ECE521 week 3: 23/26 January 2017 Outline Probabilistic interpretation of linear regression - Maximum likelihood estimation (MLE) - Maximum a posteriori (MAP) estimation Bias-variance trade-off Linear
More informationInternational Journal of Physical Sciences
Vol. 9(4), pp. 71-78, 28 February, 2014 DOI: 10.5897/IJPS2013.4078 ISSN 1992-1950 Copyright 2014 Author(s) retain the copyright of this article http://www.academicjournals.org/ijps International Journal
More informationThe multivariate skew-normal distribution
Biometrika (1996), 83, 4, pp. 715-726 Printed in Great Britain The multivariate skew-normal distribution BY A. AZZALINI AND A. DALLA VALLE Department of Statistical Sciences, University of Padua, Via S.
More informationResearch Article The Laplace Likelihood Ratio Test for Heteroscedasticity
International Mathematics and Mathematical Sciences Volume 2011, Article ID 249564, 7 pages doi:10.1155/2011/249564 Research Article The Laplace Likelihood Ratio Test for Heteroscedasticity J. Martin van
More informationTransmuted Exponentiated Gumbel Distribution (TEGD) and its Application to Water Quality Data
Transmuted Exponentiated Gumbel Distribution (TEGD) and its Application to Water Quality Data Deepshikha Deka Department of Statistics North Eastern Hill University, Shillong, India deepshikha.deka11@gmail.com
More informationWeighted Exponential Distribution and Process
Weighted Exponential Distribution and Process Jilesh V Some generalizations of exponential distribution and related time series models Thesis. Department of Statistics, University of Calicut, 200 Chapter
More informationBimodal extension based on the skew-t-normal distribution
Bimodal extension based on the skew-t-normal distribution Mehdi Amiri 1,4, Héctor W. Gómez 2, Ahad Jamalizadeh 3 and Mina Towhidi 4 Abstract: In this paper, a skew and uni-/bi-modal extension of the Student-t
More informationSome Statistical Properties of Exponentiated Weighted Weibull Distribution
Global Journal of Science Frontier Research: F Mathematics and Decision Sciences Volume 4 Issue 2 Version. Year 24 Type : Double Blind Peer Reviewed International Research Journal Publisher: Global Journals
More informationA NOTE ON GENERALIZED ALPHA-SKEW-NORMAL DISTRIBUTION. A.H. Handam Department of Mathematics Al Al-Bayt University P.O. Box , Al Mafraq, JORDAN
International Journal of Pure and Applied Mathematics Volume 74 No. 4 2012, 491-496 ISSN: 1311-8080 printed version url: http://www.ijpam.eu PA ijpam.eu A NOTE ON GENERALIZED ALPHA-SKEW-NORMAL DISTRIBUTION
More informationEstimation of Stress-Strength Reliability for Kumaraswamy Exponential Distribution Based on Upper Record Values
International Journal of Contemporary Mathematical Sciences Vol. 12, 2017, no. 2, 59-71 HIKARI Ltd, www.m-hikari.com https://doi.org/10.12988/ijcms.2017.7210 Estimation of Stress-Strength Reliability for
More informationSlash Distributions and Applications
CHAPTER 2 Slash Distributions and Alications 2.1 Introduction The concet of slash distributions was introduced by Kafadar (1988) as a heavy tailed alternative to the normal distribution. Further literature
More informationAlpha-Skew-Logistic Distribution
IOSR Journal of Mathematics (IOSR-JM) e-issn: 78-578, p-issn: 9-765X Volume 0, Issue Ver VI (Jul-Aug 0), PP 6-6 Alpha-Sew-Logistic Distribution Partha Jyoti Haaria a and Subrata Charaborty b a Department
More informationNonparametric Bayesian Methods (Gaussian Processes)
[70240413 Statistical Machine Learning, Spring, 2015] Nonparametric Bayesian Methods (Gaussian Processes) Jun Zhu dcszj@mail.tsinghua.edu.cn http://bigml.cs.tsinghua.edu.cn/~jun State Key Lab of Intelligent
More informationNew Family of the t Distributions for Modeling Semicircular Data
Communications of the Korean Statistical Society Vol. 15, No. 5, 008, pp. 667 674 New Family of the t Distributions for Modeling Semicircular Data Hyoung-Moon Kim 1) Abstract We develop new family of the
More informationMultivariate Normal-Laplace Distribution and Processes
CHAPTER 4 Multivariate Normal-Laplace Distribution and Processes The normal-laplace distribution, which results from the convolution of independent normal and Laplace random variables is introduced by
More informationAn introduction to the R package sn
An introduction to the R package sn Adelchi Azzalini Università di Padova, Italia Skewed world of data: Workshop in honor of Reinaldo B. Arellano-Valle s 65th birthday October 2017 Pontificia Universidad
More informationA Class of Weighted Weibull Distributions and Its Properties. Mervat Mahdy Ramadan [a],*
Studies in Mathematical Sciences Vol. 6, No. 1, 2013, pp. [35 45] DOI: 10.3968/j.sms.1923845220130601.1065 ISSN 1923-8444 [Print] ISSN 1923-8452 [Online] www.cscanada.net www.cscanada.org A Class of Weighted
More informationarxiv: v3 [stat.me] 25 Aug 2017
Natural (non-)informative priors for skew-symmetric distributions Holger Dette a, Christophe Ley b, and Francisco J. Rubio c a Ruhr-Universität Bochum, Germany. arxiv:1605.02880v3 [stat.me] 25 Aug 2017
More informationIntroduction of Shape/Skewness Parameter(s) in a Probability Distribution
Journal of Probability and Statistical Science 7(2), 153-171, Aug. 2009 Introduction of Shape/Skewness Parameter(s) in a Probability Distribution Rameshwar D. Gupta University of New Brunswick Debasis
More informationFitting age-specific fertility rates by a skew-symmetric probability density function
Fitting -specific fertility rates by a skew-symmetric probability density function Stefano Mazzuco Bruno Scarpa Department of Statistical Sciences, University of Padova Via Cesare Battisti, 241 35121,
More informationThe Use of Copulas to Model Conditional Expectation for Multivariate Data
The Use of Copulas to Model Conditional Expectation for Multivariate Data Kääri, Meelis; Selart, Anne; Kääri, Ene University of Tartu, Institute of Mathematical Statistics J. Liivi 2 50409 Tartu, Estonia
More informationA Constructive Representation of Univariate Skewed Distributions
A Constructive Representation of Univariate Skewed Distributions José T.A.S. Ferreira and Mark F.J. Steel Department of Statistics University of Warwick, UK Abstract We introduce a general perspective
More informationStat 451 Lecture Notes Markov Chain Monte Carlo. Ryan Martin UIC
Stat 451 Lecture Notes 07 12 Markov Chain Monte Carlo Ryan Martin UIC www.math.uic.edu/~rgmartin 1 Based on Chapters 8 9 in Givens & Hoeting, Chapters 25 27 in Lange 2 Updated: April 4, 2016 1 / 42 Outline
More information