ANALISIS SUPPORT VECTOR REGRESSION (SVR) DALAM MEMPREDIKSI KURS RUPIAH TERHADAP DOLLAR AMERIKA SERIKAT
In economy, the global markets have an important role as a forum for international transactions between countries in selling or purchasing goods or services on an international scale. Money as legal tender in the trading activities, but the problem is the difference between the state of the currency...
Saved in:
| Main Author: | |
|---|---|
| Format: | Academic Paper |
| Published: |
2014-10-30.
|
| Subjects: | |
| Online Access: | http://eprints.undip.ac.id/46349/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In economy, the global markets have an important role as a forum for international transactions between countries in selling or purchasing goods or services on an international scale. Money as legal tender in the trading activities, but the problem is the difference between the state of the currency, the exchange rate will be established. Exchange rate is the value of a country's currency is expressed in another country's currency value. Fluctuations in foreign exchange rates greatly affect the Indonesian economy, so the determination of the exchange rate should be beneficial to a country can run the economy well. To predict the exchange rate of the Rupiah against the United States dollar in this study used methods of Support Vector Regression (SVR) is a technique to predict the output in the form of continuous data. SVR aims to find a hyperplane (line separator) in the form of the best regression function is used to predict the exchange rate against the United States dollar with linear kernel and polynomial functions. Criteria used in measuring the goodness of the model is the MAPE (Mean Absolute Percentage Error) and R2 (coefficient of determination). The results of this study indicate that both the kernel function gives very good accuracy in the prediction results of the exchange rate with R2 of 99.99% with MAPE 0.6131% in the kernel linear and R2 result of 99.99% with MAPE 0.6135% in the kernel polynomial. Keyword : Exchange rate, Support Vector Regression (SVR), Hyperplane, Linear Kernel, Polynomial Kernel, ε-insensitive, Accuracy |
|---|---|
| Item Description: | http://eprints.undip.ac.id/46349/1/RIZKY_AMANDA.pdf |