1 AIT Asian Institute of Technology

A simulation study on order determination of ARMA models

AuthorChandrabalan, Kanagasabai
Call NumberAIT SSPR no.CA-81-1
Subject(s)Time-series analysis--Computer programs

NoteA special study submitted in partial fulfillment of the requirements of the degree of Master of Engineering
PublisherAsian Institute of Technology
AbstractThis study is concerned with tuilding a suitable autoregresSive— moving average model, which posseSSes the maximum singlicity and the minimum number cf parameters consonant With representational adequacy, rcr discrete time series. An attempt is made in this study to consider the order or an autoregreSSive—moving average model selected to reprESent ; set of data, determined by some important methods. USing simulated data from several useful models, it is found that the modified portmanteau statistic is least sensitive with regard to model misspecification while the Akaike Internaticn Criterion (AIC) and mcoifieo Akaike information Criterion (51C) give nearly the same results in most cases. Ozaki,s minimum criteria (01C) always gives the same result as AIC. Annual streamflow SEQUence can be modeled by low order autoregressive processes.
Year1981
TypeSpecial Study Project Report (SSPR)
SchoolSchool of Engineering and Technology (SET)
DepartmentDepartment of Information and Communications Technologies (DICT)
Academic Program/FoSComputer Application (CA)
Chairperson(s)Huynh Ngoc Phien;
Examination Committee(s)Hosking, Roger J. ;Boonserm Weesakul;
Scholarship Donor(s)Government of Norway;
DegreeSpecial Studies Project Report (M. Eng.) - Asian Institute of Technology, 1981


Usage Metrics
View Detail0
Read PDF0
Download PDF0