Options
Time series regression model : an application to Singapore gross domestic product series
Author
Yong, Hua Moy
Supervisor
Cheang, Wai Kwong
Abstract
This academic exercise uses time series regression model to analyse time series data which have the distinctive feature of being auto-correlated. The theory is applied to the Singapore Annual Gross Domestic Product (GDP) series. By identifying a suitable regression model for the GDP series and an autoregressive (AR) mode for the noise term, we are able to make predictions and inferences about the data. We also show through simulation that the restricted maximum likelihood (REML) approach produces a less biased estimate for the noise parameter in the time series regression model with autoregressive AR(1) noise than the maximum likelihood (ML) estimation. As a result, we also achieve better estimates for the regression parameters.
Date Issued
2003
Call Number
QA280 Yon
Date Submitted
2003