|
Activity Number:
|
150
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #308255 |
|
Title:
|
An Empirical Spectral Test (EST) for Random Sequences
|
|
Author(s):
|
David Zeitler*+ and Joseph W. McKean and John Kapenga
|
|
Companies:
|
Grand Valley State University and Western Michigan University and Western Michigan University
|
|
Address:
|
1 Campus Drive, Allendale, MI, 49401,
|
|
Keywords:
|
empirical spectral test ; random number generator ; random sequence test ; Fourier transform
|
|
Abstract:
|
Random number sequences are critical to research in many fields. The validity of research results often depend heavily on the underlying distribution of these sequences. In this work we introduce the Empirical Spectral Test (EST). The EST is a highly flexible test of spatial uniformity based on a multi-dimensional Fourier transform of the empirical probability density function. It has properties in common with theoretical tests such as the spectral test and discrepancy tests. However the EST can be applied to sequences from any random number source, can be adapted to specific user requirements and has the added advantage that its computational complexity is relatively independent of the number of data points being tested. This later makes it particularly interesting as a test for extremely long period generators used for parallel computation work.
|
