<![CDATA[The management of water resources and various industrial sectors is highly dependent on rainfall. To avoid negative impacts such as floods, droughts and other natural disasters, rainfall forecasts must be accurate and timely.This research aims to find the best algorithm for predicting rainfall. In this study, modeling was carried out using the Bandung city rainfall dataset from 2018 to 2022 using the Bidirectional Gated Recurrent Unit (BiGRU) method. Bidirectional Long Short Term Memory (BiLSTM), Gated Recurrent Unit (GRU), Long Short Term Memory (LSTM) are used to compare the performance of the BiGRU algorithm. The test findings show that, with value Root Mean Squared Error (RMSE) and R2 Score BiGRU gives the best results with the lowest error rate. The algorithm with the biggest error rate is LSTM. This study advances strategies for predicting rainfall that can be applied to managing water resources and responding to natural disasters related to rainfall.]]>
Copyrights © 2023
