The main objective of this study is to describe the simulation of data that contain the four time series components in univariate forecasting. Time series data occur in a mixture of all the four time series components in its natural form (trend component, seasonal components, cyclical components, irregular components). Simulated data are generated in mixture of all the components before separating each components to its individual group. Statistical techniques is being required in the traditional system to separate those components into its various individual components to archive smooth forecast. This study provide an alternative to the conventional method of decomposition of time series data. This bridge the gap between the expert and end users in search for techniques that would be used in decomposing time series data. The simulation of data with all the basic time series components also reduced time and helps in getting the exact proportion of the time series components needed for any further analysis. Three simulation data size was carried out. The trend data was generated through a regression line and other subsequent time series data was manually added(Box, Jenkins, Reinsel, & Ljung, 2015). The new approach reduce time of decomposing time series data, easy to conduct, easy to acquire, can be easily controlled, efficient and reliable, hence this approach is recommended for any statistical analysis that required time series data for univariate forecasting.
Volume 11 | 05-Special Issue
Pages: 1005-1010