Perbandingan Akurasi ARIMA dan Backpropogation dalam Memprediksi Intensitas Curah Hujan Kota Makassar

Andi Illa Erviani Nensi (Universitas Negeri Makassar)

Integrated Lab Journal
Vol. 11 No. 01 (2023): Integrated Lab

Climatic conditions tend to change irregularly in a random period of time, so that unavoidable deviations can occur such as high intensity of rainfall which causes flooding. This has an impact on people who feel disadvantaged because floods can paralyze economic activities, threaten public safety, hamper transportation and the emergence of other negative impacts, especially in Makassar City. Therefore, it is important to do a rainfall forecast to find out information about rainfall in Makassar City so that in this study a seasonal rainfall prediction was carried out by applying the Auto Regressive Integrated Moving Average (ARIMA) and Backpropogation methods. The results of the ARIMA method with the model produced the smallest error of 174,9038 while the backpropogation method with the 12-18-18-18-1 model gave the smallest error of 199, 0376. This shows that the ARIMA method has a better level of accuracy in forecasting rainfall compared to the backpropogation method.