Accession Number : ADA186427

Title :   On the Characterization of Certain Point Processes.

Descriptive Note : Technical rept. Sep 86-Aug 87,

Corporate Author : NORTH CAROLINA UNIV AT CHAPEL HILL DEPT OF STATISTICS

Personal Author(s) : Hsing, Tailen

PDF Url : ADA186427

Report Date : Aug 1987

Pagination or Media Count : 23

Abstract : It is well known that point process methods can be applied effectively to study certain types of problems in statistical extreme value theory. Consider a strictly stationary sequence of random variables (xi sub j) indexed by the set of integers I=Z. One can define a number of interesting point processes in one dimension by recording the positions where extreme values occur. For example, an extremal process typically is one that records the indices (properly normalized) at which record values of xi sub 1, xi or sub 2 occur, and an exceedance point process considered by Leadbetter consists of the set of points j.n: xi sub j w sub n, where sub n is a suitable sequence of constants. For this type of processes, Poisson or compound Poisson convergence results can often be derived under suitable mixing conditions. Keywords: Weak convergence.

Descriptors :   *POISSON DENSITY FUNCTIONS, *WEAK CONVERGENCE, *POINT THEOREM, CONSTANTS, CONVERGENCE, MIXING, RANDOM VARIABLES, RANGE(EXTREMES), SEQUENCES, STATIONARY, STATISTICS, THEORY, VALUE

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE