Foreign exchange rate prediction between indonesian rupiah and us dollar using transductive learning
Transductive learning which is focused on creating the local or individual model is expected to have the capability to solve the exchange rate prediction problem, especially for the fluctuation of the exchange rate. Therefore, as the aim of this study, new algorithm is proposed to evaluate the capability of transductive learning in the case of foreign exchange prediction. This study developed an automated system to predict the exchange rate using proposed method called Polynomial Regression based Transductive Learning. It also implements the similar pattern retrieval in the term of generating the predicted value. The proposed method was evaluated using exchange rate data between the Indonesian rupiah and the U.S. dollar which spans 24th January 2001 to 31th May 2009. The experimental results showed that the proposed method has the capability to provide a high accuracy at average, and it can also retrieve the similar pattern from past.
B00542 | (wh) | Available |
No other version available