DefinePK

Abstract

Semarang is an administrative city in Central Java province that is inevitably suffer from tidal flooding phenomenon. Tidal flooding is caused by the rising of sea level. Forecasting methods are techniques in Statistical tools for decision making. Therefore, a forecasting of sea level becomes important. One of the method to forecast time series data is ARIMA which require fulfillment of assumptions. One other way to put aside assumptions is by using ARIMAX. Meanwhile, non-linear approach that does not require assumptions fulfillment is ANFIS. The forecasting of sea level using ARIMAX is better than ARIMA and ANFIS. It shows that a certain complex forecasting methods does not guarantee to result the best model. The resulting model is ARIMAX (0, 1, [3]) (1, 0, 0)12 with 7 outliers which produces 4.82 of RMSE.