Implementation of α-Sutte Indicator to Forecasting Consumer Price Index in Turkey
| Main Authors: | Ahmar, Ansari Saleh, Rahman, Abdul, Mulbar, Usman |
|---|---|
| Format: | Proceeding NonPeerReviewed Book |
| Bahasa: | eng |
| Terbitan: |
, 2017
|
| Subjects: | |
| Online Access: |
http://eprints.unm.ac.id/2851/1/IConMNS2017-Sutte%20Indicator.pdf http://eprints.unm.ac.id/2851/ http://fmipa.undiksha.ac.id/iconmns2017/ |
Daftar Isi:
- α-Sutte Indicator (α-Sutte) was originally from developed of Sutte Indicator. Sutte Indicator can use to predict the movement of stocks. As the development of science, then Sutte Indicator developed to predict not only the movement of stocks but also can forecast data on financial, insurance, and time series data. This method called α-Sutte Indicator (α-Sutte). α-Sutte was developed using the principle of the forecasting method of using the previous data. The data used in this research is Consumer Price Index in Turkey data from January 2003 - June 2017. This data is divided into 2 parts, namely training data and test data. Training data starts from January 2003 - October 2016 and test data from November 2016 - June 2017. To see the accuracy of α-Sutte, it will be done benchmarking the results of forecasting with other forecasting method is Automatic Time Series Forecasting: The forecast Package for R (AutoARIMA) developed by Hyndman-Khandakar (2008). Comparison of this accuracy is to compare the value of MSE forecasting result on test data by using training data as reference data. Results obtained from this study that the MSE value of α-Sutte is smaller (5.697723) than MSE from AutoARIMA (292.5125). This indicates that α-Sutte is more suitable for predicting Consumer Price Index in Turkey data.