PENERAPAN METODE ANFIS UNTUK MEMPREDIKSI NILAI TUKAR DOLAR AS (USD) TERHADAP RUPIAH

Abstract

The purpose of this study is to explain the prediction of the exchange rate of the United States Dollar (USD) on the Indonesia Rupiah (IDR) using the Adaptive Neuro Fuzzy Inference System (ANFIS) method. ANFIS is a combination of functions contained in fuzzy logic and neural networks. The data used is data with daily period of the USD exchange rate against the rupiah. The input to the daily prediction is a combination of the exchange rates on the previous days for the prediction of the selling rate of the United States Dollar (USD) for the next one day. The results of this study, the best model ANFIS uses a membership function Gaussian with model 3 consisting of 4 inputs and 1 output with a combination of training data and test data 80:20 or 202 data training and 51 data testing resulting MAPE is 3.74% and from the MAPE value it can be concluded that predictions using the ANFIS can be categorized as very accurate.