Prediksi Awal Musim Hujan Menggunakan Adaptive Neuro-Fuzzy Inference System pada Studi Kasus Kabupaten Indramayu

Abstract

Climate anomalies cause the onset of the rainy season changes in Indonesia. In agriculture, the change complicates precise determination of cultivation. Incorrect determination increases crop failure due to droughtness. As one of the national granaries, rice production in Indramayu highly depends on the climate. This research aims to predict the onset of the rainy season in Indramayu using adaptive neuro-fuzzy inference system (ANFIS). The data used in this study are southern oscillation index (SOI) from Mei to August as predictor and the onset of the rainy season from 1971 to 2010. SOI data was reduced to 7 intervals. There are 5 rain regions in Indaramayu, and the average rain region was predicted using ANFIS model for each region. The prediction result was evaluated using the root mean squared error and the squared correlation coefficient. The best prediction result was obtained at interval [-7,7] with an average of the root mean squared error of 19.5 days and a squared coefficient correlation of 0.33.