PREDIKSI CURAH HUJAN BULANAN MENGGUNAKAN TIME SERIES (SINGLE EXPONENTIAL SMOOTHING) DAN KNN (STUDI KASUS : KABUPATEN PADANG PARIAMAN)

Harsani, Prihastuti
Mulyana, Iyan
Ofik Hidayat, Ade

Proceedings of KNASTIK
2012

Climate is average condition for certain time period. The one of few factors influencethe climate is rainfall. This research develops a system which can predict and giveinformation of monthly rainfall prediction and also for the next few months.Climatology data from BMKG from 2000 to 2009 is need in this research. The dataconsist of seven variable which are temperature, sunlight condition, air pressure,humidity, wind speed, wind direction and evaporate. This research uses ExponentialSmoothing for rainfall prediction and K-Nearest Neighbor that need for rainfallprediction classified. The result shows that the prediction accuracy is 58, 33%.