Page 223 - Contributed Paper Session (CPS) - Volume 2

CPS1834 Gumgum D. et al.
calendar. In the second step we have two time series data, rainfall data based
on Hijri and Gregorian. In the third step, both time series data are modelled
by SARIMA model.
Sarima modeling was carried out for both types of data, namely Gregorian
and Hijri. Sarima modeling is done by tracking, the Sarima model that matches
/ matches the two data. The Next step, determination of MAD,MSE,MAPE and
AIC of all models. This best model is chosen if it meets several criteria, the
parameters are significant, Residual is white noise (Independent and follows
Normal distribution). The fifth step is comparison of all result in various
horizon forecasting, the last step is conclusion. Steps in this methodology can
be drawn in the flow chart below

Rainfall Daily Data

monthly rainfall data monthly rainfall data
(Gregorian) (Hijriya)

SARIMA Modeling SARIMA Modeling

Determine MAD,MSE Determine MAD,MSE
MAPE and AIC. MAPE and AIC.

Comparison

Result

Figure 1. Methodology Step of SARIMA Modelling Based on Hijri and Gregorian
Calendar

212 | I S I W S C 2 0 1 9