<![CDATA[The aim of this research is to provide the results of ARIMA modeling on rainfall data in Langsa City in 2017-2021. The initial stage of ARIMA modeling is the identification of data stationarity. Meanwhile, stationarity in the mean can be done with data plots and ACF forms. Identification of ACF and PACF forms from data that is already stationary is used to determine the order of the alleged ARIMA model. The next stage is parameter estimation to see the suitability of the model. The diagnostic check process is carried out to evaluate whether the residual model meets the white noise requirements and is normally distributed. The Ljung-Box test is a test that can be used to validate white noise requirements. Rainfall data forms a stationary time series. Furthermore, from the model fit test it was found that the MA(1) model was suitable for predicting the model. Meanwhile, AR(1) and ARMA(1,1) are not used to predict because they do not meet the model fit test. The model obtained with the MA(1) model is as follows, namely .]]>
Copyrights © 2022
