Jurnal Info Sains : Informatika dan Sains
Vol. 13 No. 01 (2023): Jurnal Info Sains : Informatika dan Sains , Maret 2023

RAINFALL FORECASTING USING THE HOLT-WINTERS EXPONENTIAL SMOOTHING METHOD

I Komang Arya Ganda Wiguna (Fakultas Teknologi Dan Informatika , Institut Bisnis Dan Teknologi Indonesia)
Ni Luh Putu Ayu Cintia Utami (Fakultas Teknologi Dan Informatika , Institut Bisnis Dan Teknologi Indonesia)
Wayan Gede Suka Parwita (Fakultas Teknologi Dan Informatika , Institut Bisnis Dan Teknologi Indonesia)
I Putu Agus Eka Darma Udayana (Fakultas Teknologi Dan Informatika , Institut Bisnis Dan Teknologi Indonesia)
I Gede Iwan Sudipa (Fakultas Teknologi Dan Informatika , Institut Bisnis Dan Teknologi Indonesia)



Article Info

Publish Date
29 Mar 2023

Abstract

Abiansemal District is dominated by agricultural land, with Subak Blahkiuh being one of the agricultural lands in the district. Agriculture is extremely weather-dependent, particularly during the monsoon season. Therefore, precipitation forecasting is necessary for determining a favorable sowing season. Researchers use the Holt-Winters Exponential Smoothing method with data from 2012 to 2022 for forecasting. Holt-Winters Exponential Smoothing employs two forecasting models. The results of the calculations indicate that the additive model has alpha equal to 0.653467, beta equal to 0.0036348, and gamma equal to 0.1182400. While the multiplicative model has alpha values of 0.8889286, 0.0001 and 0.0246825 for beta and gamma, respectively. For accurate forecasting, it is necessary to examine the error data, specifically the MAE and MAPE. The multiplicative model yielded an MAE of 158.87 and a MAPE of 55.18%, whereas the additive model yielded an MAE of 186.59 and a MAPE of 70.18%. The MAPE values of the additive model and multiplicative model are greater than 50 percent, indicating that the additive and multiplicative models provide inaccurate forecasts. Between the additive model and the multiplicative model, the Holt-Winters Exponential Smoothing method favors the multiplicative model for future precipitation forecasting.

Copyrights © 2023






Journal Info

Abbrev

InfoSains

Publisher

Subject

Computer Science & IT

Description

urnal Info Sains : Informatika dan Sains (JIS) discusses science in the field of Informatics and Science, as a forum for expressing results both conceptually and technically related to informatics science. The main topics developed include: Cryptography Steganography Artificial Intelligence ...