Premium
An evaluation of test statistics for detecting temporary change in Bl(1,1,1,1) models
Author(s) -
Zaharim Azami,
Ismail Amiruddin,
Abdullah Shahrum,
Mohamed Ibrahim,
Ahmad Ibrahim
Publication year - 2007
Publication title -
pamm
Language(s) - English
Resource type - Journals
ISSN - 1617-7061
DOI - 10.1002/pamm.200700953
Subject(s) - outlier , statistics , anomaly detection , sampling (signal processing) , sample (material) , sample size determination , mathematics , change detection , computer science , data mining , artificial intelligence , chemistry , chromatography , detector , telecommunications
A study is carried out to investigate the sampling properties of the outlier test statistics of a procedure developed for detecting temporary change in BL(1,1,1,1) processes. It is done with respect to the sample size, the type of outlier and the size of the coefficients of the BL(1,1,1,1) process. The results show that, in general, the outlier detection procedure is capable of detecting TC, although the performance is affected if ω is large. (© 2008 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)