Application of the Attention-Based LSTM Method for Rainfall Prediction in East Java

This research aims to measure the performance of the Attention-Based Long Short-Term Memory (LSTM) predictive model in rainfall prediction analysis in East Java, with a focus on including the application of the model in predicting complex time-series data. The main objective of this study is to create an efficient and accurate model and to emit the performance of the Attention-Based LSTM algorithm compared to conventional methods. The methodology used includes rainfall data collection, data preprocessing, Attention-Based LSTM model design, training models, and testing to assess accuracy. The results of the study indicate that the Attention-Based LSTM model is able to improve rainfall prediction compared to conventional methods, with the Root Mean Squared Error (RMSE) evaluation metrics with a value of 0.00807 and Mean Squared Error (MSE) with a value of 0.08987 which shows better results, so this model can be relied on for real-world applications.