1 AIT Asian Institute of Technology

Estimation of parameters in regression models with correlated disturbances by the method of maximum likelihood

AuthorFong, Lye Mui
Call NumberAIT Thesis no. CA-85-9
Subject(s)Regression analysis--Computer programs
NoteA thesis submitted in partial fulfillment of the requirements for the degree of Master of Science, School of Engineering and Technology
PublisherAsian Institute of Technology
AbstractEstimation of parameters in linear regression models with disturbances following AR (p) or MA(q) processes has been common. But estimation of models with ARMA(p, q) errors has only been considered theoretically. This thesis is concerned with the software development of a general linear regression model with correlated disturbances that follow an ARMA (p, q) process. This also includes estimation of models with AR or MA errors. The methodology was based on the paper of Harvey and Phillips (1979) , whereby the system design and program were developed accordingly. Applications to simulated data and actual data were used to test the efficiency and workability of the software developed. This thesis showed that the method of exact maximum likelihood, together with Kalman filter and recursive residuals, does in fact produce good estimators with relative errors < 20%, and is found a better estimation method than others.
Year1985
TypeThesis
SchoolSchool of Engineering and Technology (SET)
DepartmentOther Field of Studies (No Department)
Academic Program/FoSComputer Application (CA)
Chairperson(s)Huynh, Ngoc Phien
Examination Committee(s)Clarke Harry R. ;Vilas Wuwongse
Scholarship Donor(s)The Government of Germany
DegreeThesis (M.Sc.) - Asian Institute of Technology, 1985


Usage Metrics
View Detail0
Read PDF0
Download PDF0