This document discusses the mathematical theory of risk and explores the convergence of maximum values from different distributions as n goes to infinity. It covers the Fisher-Tippett-Gnedenko Theorem, the three classes of extreme value distributions, and the simulation of samples for fitting the GEV distribution.
Contribute Materials
Your contribution can guide someone’s learning journey. Share your
documents today.
ACST356 / ACST861 Mathematical Theory of Risk1 ACST356 / ACST861 Mathematical Theory of Risk 2019 Assignment 2 By (Name of Student) (Institutional Affiliation) (Date of Submission) 1
Secure Best Marks with AI Grader
Need help grading? Try our AI Grader for instant feedback on your assignments.
ACST356 / ACST861 Mathematical Theory of Risk2 Part A Question 1(extreme value theory) Considernrandom variablesX1,X2, … ,Xn, which are independent and identically distributed (iid) with cumulative distribution function (cdf)F x( ). Their maximum value is expressed asMnmax(X X1,2, ... ,Xn) . The Fisher-Tippett-Gnedenko Theorem states that under certain technical conditions, there exist a sequence of positive numbersanand a sequence of numbersbnsuch that asngoes to infinity, lim Pr(Mnbnx)G x( ) , nan in whichG x( )is the cdf of the generalised extreme value (GEV) distribution, denoted as GEV( , , )forand0. The limiting cdfG x( )must come from one of the three classes: Gumbel (0), Fréchet (0), and reversed Weibull (0), which are given x G x( ) exp( (1) ), 0,x, x G x( )exp( (1) ),0,x. Deriving from first principles, find out to which class a maximum value from the (a) Exponential(1), (b) Pareto(1,1) (Type I), and (c) Uniform(0,1) distribution converges asn goes to infinity. Identify clearly the values of,, andunder each underlying distribution. SupposeX1,X2,...,are independent random variables with the same probability distribution, and letMn= max(X1,...,Xn). Under certain circumstances, it can be shown that there existnormalizing constantsan>0,bnsuch that Pr. 2 below: x G x( )exp( exp()), 1 0,x, 1
ACST356 / ACST861 Mathematical Theory of Risk3 TheThree Types Theorem(Fisher-Tippett, Gnedenko) asserts that if nondegenerateH exists, it must be one of three types: H(x) = exp(−e−x),allx(Gumbel) (Fr´echet) (Weibull) In Fr´echet and Weibull,>α0. The three types may be combined into a singlegeneralized extreme value(GEV) distribution: , (y+= max(y,0)) whereμis a location parameter,>ψ0 is a scale parameter andξis a shape parameter.ξ→0 corresponds to the Gumbel distribution,>ξ0 to the Fr´echet distribution withα= 1/ξ,<ξ0 to the Weibull distribution withα=−1/ξ.>ξ 0: “long-tailed” case, 1−F(x)∝x−1/ξ, ξ= 0: “exponential tail” <ξ0: “short-tailed” case, finite endpoint atμ−/ξ ψ Question 2(simulation) Referring to Question 1 above, simulate 1,000 samples ofM1,000from the (a) Exponential(λ), (b) Pareto(α,β) (Type I), and (c) Uniform(a,b) distribution. Choose your own parameter values that are different to those in Question 1. Use maximum likelihood to fit the GEV distribution to the simulated samples. Check the goodness-of-fit and also explain the results under the context of extreme value theory. 3
ACST356 / ACST861 Mathematical Theory of Risk4 References Castillo, E. (2012).Extreme value theory in engineering. Elsevier. De Haan, L., & Ferreira, A. (2017).Extreme value theory: an introduction. Springer Science & Business Media. Galambos, J. (2014). Extreme value theory for applications. InExtreme value theory and applications(pp. 1-14). Springer, Boston, MA. Kotz, S., & Nadarajah, S. (2010).Extreme value distributions: theory and applications. World Scientific. 4