The following information was submitted:
Transactions: WSEAS TRANSACTIONS ON MATHEMATICS
Transactions ID Number: 32-761
Full Name: Raden Mohamad Atok
Position: Ph.D. Candidate
Age: ON
Sex: Male
Address: EF33 Lorong Aman 1A, Taman Kajang Utama, 43000, Selangor
Country: MALAYSIA
Tel: +603-8921-6843
Tel prefix: +603
Fax: +6013-89216960
E-mail address: atok@vlsi.eng.ukm.my
Other E-mails: atok.stat@yahoo.com
Title of the Paper: A Simulation Study of Single Outlier Detection for Temporary Change in ARMA (1, 1) Data
Authors as they appear in the Paper: Azami Zaharim, Raden Mohamad Atok, Rafizah Rajali, Dzuraidah Abdul Wahab
Email addresses of all the authors: azami@eng.ukm.my, atok@eng.ukm.my, rafizah_rajali@yahoo.com, dzuraida@eng.ukm.my
Number of paper pages: 10
Abstract: The aim of this research is to carry out appropriate method to detect Temporary Change in ARMA (1, 1) data. Estimation of model parameters and outlier effects are used to iteratively for joint estimation procedure. The data simulation is generated from ARMA (1,1) model using 4 parameters combination of AR and MA. Four different approaches of variance calculation are used to detect presence of outlier, i.e. Ordinary Least Square (LS) using Omit-one data, LS with replacing data by fit, Median Absolute Deviation (MAD) using Omit-one data and MAD with replacing data by fit. The result shows Omit-one approach can detect outlier better than Fit approach for all cases. Furthermore MAD-Omit-one is slightly better than LS-Omit-one. The result above is implemented to Surabaya's Air Pollutant data get same result, MAD-Omit one is slightly more powerful to detect outlier. Conclusion of this research is Joint Estimation method using MAD-Omit-one as approach of variance calcula!
tion give better result than three other approaches.
Keywords: ARMA, Joint Estimation, Median Absolute Deviation, Omit-one, Simulation, Single Outlier Detection, Temporary Change
EXTENSION of the file: .doc
Special (Invited) Session:
Organizer of the Session:
How Did you learn about congress:
IP ADDRESS: 202.185.32.240
