Forecasting and evaluation of time series with multiple seasonal component
Seasonality is one of the components in time series analysis and this seasonal component may occur more than one time. Thus, modelling the seasonality by using one seasonal component is not enough and could produce less forecast accuracy. Autoregressive Integrated Moving Average (ARIMA) models is th...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Universiti Putra Malaysia
2021
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/97378/ http://psasir.upm.edu.my/id/eprint/97378/1/13922-2190-46665-1-10-20210619.pdf |
| Summary: | Seasonality is one of the components in time series analysis and this seasonal component may occur more than one time. Thus, modelling the seasonality by using one seasonal component is not enough and could produce less forecast accuracy. Autoregressive Integrated Moving Average (ARIMA) models is the fundamental method in developing the seasonal ARIMA for one seasonality or more than one seasonality. Therefore, to validate the method performance, the hourly air quality data with double seasonality were carried out as the case study. The model identification step to determine the order of ARIMA model was done by using MINITAB program and the model estimation step by using SAS program and Excel. The results showed that the double seasonal ARIMA able to model and forecast the air quality data with high frequency. |
|---|