Procedure for Estimation of Additive Time Series Model
- April 8, 2021
- Posted by: RSIS
- Categories: IJRSI, Statistics
International Journal of Research and Scientific Innovation (IJRSI) | Volume VIII, Issue II, February 2021 | ISSN 2321–2705
Procedure for Estimation of Additive Time Series Model
K.C.N. Dozie1, M.U. Uwaezuoke2
1Department of Statistics Imo State University, Owerri, Imo State, Nigeria
2Department of Mathematics Imo State University, Owerri, Imo State, Nigeria
Abstract: The procedure for estimation of lineartrend cycle and seasonal components and accepts additive model is examined in this study. Estimates of the periodic, seasonal and overall means and variances with error terms and error variances are obtained for additive model. Empirical example based on short series in which trend cycle component is jointly estimated for the linear case is applied to determine suitable model for decomposition of the study series.
Keywords: Descriptive Time Series, Additive Model, Error Term, Buys-Ballot Estimate, Error Variance, Suitable Model.
I. INTRODUCTION
Descriptive method involves the assessment of trend, seasonality, cycles, changes in level, turning points and so on that may influence the series. The reason of time series decomposition method is to separate the time series components available in the series. The components are; i) the trend component ii) the seasonal component iii) the cyclical component iv) the irregular component. The trend and cyclical component could be estimated to obtain the trend-cycle component in short time period [1].Therefore, the decomposition models are
