PERAMALAN HARGA PARITAS KEDELAI MODEL ANFIS

Azis Muslim (Badan Pengkajian dan Pengembangan Kebijakan Perdagangan Kementerian Perdagangan Jalan M. I. Ridwan Rais No. 5, Jakarta Pusat.)

Widyariset
Vol 17, No 1 (2014): Widyariset

The aim of this research is to compare the accuracy of the ARMA conventional model and the ANFIS softcomputing model in forecasting soybean parity price. Two variables are chosen, Rupiah exchange rate and international soybean price which are the most volatile than others that make up the parity price. The methods used to compare the two models are Theil’s Inequality and MAPE. The results show that the ANFIS models are superior in predicting the exchange rate and international soybean prices based on the selected performance method.