<![CDATA[Abstract: The climate anomaly that occurs is La Nina conditions, this condition is in the central part of the Pacific Ocean Sea Surface Temperature (SML) has cooled below normal conditions. The impact of La Nina conditions is the occurrence of heavy rains that cause flood disasters. This study aims to predict areas prone to flooding in West Kalimantan based on rainfall data using training data in 2019-2021 testing data in 2022 with a ratio of 80:20 and test the accuracy and RMSE value using the H2O deep learning algorithm in the hidden layer 100,50,100. This study uses climate data from 5 climatology stations in Sintang, Melawi, Kapuas Hulu, Sambas and Ketapang districts. The results obtained in this study are by conducting experiments for 10 times using the number of epochs 10, 15, 20, 25, 30, 35, 40, 45, 50, and 55 resulting in the best accuracy and the lowest RMSE value at the number of epochs 30 with an accuracy of 99.54% and RMSE worth 0.087. So it can be concluded that the deep learning algorithm is able to predict flood disasters well based on rainfall data with the best accuracy value is 90.54% and the lowest RMSE is 0.087. ​Keywords: Floods, Rainfall, Deep Learning H2O, Prediction]]>
Copyrights © 2023
