PREDIKSI JANGKA PENDEK BULANAN JUMLAH FLARE DENGAN MODEL ARIMA (p,d,[q]), (P,D,Q)132

Abstract

The time series of the monthly of flares from January 1965 to December 2004 is analyzed by ARIMA models. It is suggested that there is any auto correlation between monthly data in regular and seasonal order (11 years solat activity cycle) in the ARIMA (p,d[q]), (P,D,Q)132 models. After several of feasibility test had successed and the assumption of stationary in mean and variance was obtained, we can determine the model i.e; ARIMA (1,1,[13])(0,1,1)132 and ARIMA (0,1,[13])(0,1,1)132. From this models, we can use to predict the monthly number of flares.