Comparison Of Jenkins Box Method And Multiple Linier Regression In Predicting The Noble Metal Price
| Main Authors: | Gunara, Bayu, Yennimar , Yennimar |
|---|---|
| Format: | Article info application/pdf eJournal |
| Bahasa: | eng |
| Terbitan: |
Institute of Computer Science (IOCS)
, 2019
|
| Subjects: | |
| Online Access: |
https://medikom.iocspublisher.org/index.php/JTI/article/view/15 https://medikom.iocspublisher.org/index.php/JTI/article/view/15/15 |
Daftar Isi:
- Nowadays, gold investment practitioners generally use instinct and guess in investing in gold. This is certainly a problem because it has a high error margin. To solve these problems, the forecasting process can be carried out.To be able to forecast gold prices with low error rates, various studies have been conducted. The Box-Jenkins method performs better than other methods in predicting the price of gold, because the Box-Jenkins method applies forecasting by relying on the historical statistics of gold prices beforehand. The Box-Jenkins method is an iterative of choosing the best model for the stationary series of a group of linear time series models called the ARIMA (Autoregressive Integrated Moving Average) model. However, the ARIMA method is a complex method and is not easy to use and requires a long execution time to obtain forecasting results with a high degree of accuracy. To improve the accuracy of prediction results from ARIMA, the ARIMA method can be combined with the multiple regression method into a hybrid method. The Multiple Linear Regression Method is a mathematical technique that minimizes the difference between the actual value and the predicted value.The results of this study are an application of forecasting the price of gold using the ARIMA method and Multiple Linear Regression. The application also provides a facility to test the results of the methods used. Based on the results of testing the accuracy of the prediction results from the hybrid method with 30 data = 48%, 60 data = 40%, and 118 data = 40.81%.