Prediksi Nilai Tukar Rupiah Indonesia Terhadap Dolar Amerika Serikat Menggunakan Metode Recurrent Extreme Learning Machine Neural Network

Daneswara Jauhari (Fakultas Ilmu Komputer, Universitas Brawijaya)
Imam Cholissodin (Fakultas Ilmu Komputer, Universitas Brawijaya)
Candra Dewi (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 1 No 11 (2017): November 2017

The exchange rate of money by some people who are involved in the economy, especially the inter-state economy is very payed, often influencing one's decision in taking a policy. However, the exchange rate is a very unstable value, has a lot of noise and fluctuation, it is very difficult to predict the exchange rate. Research on exchange rate prediction has become the most challenging research among researchers, and that is considered one of the most important areas of research in international finance. Therefore, an application is needed, which can better predict the exchange rate of Indonesian Rupiah against the US Dollar. In this study the authors use the method of Recurrent Extreme Learning Machine Neural Network (RELMNN), the method can handle time-ordered datasets and can improve the ability of the Extreme Learning Machine (ELM) method in training and adapting. After testing with optimum parameters, and compared with ELM method, we found out that RELMNN method is superior to ELM method with Mean Absolute Percentage Error (MAPE) value of 0.069502%, while ELM method get MAPE 0.090423%.