Increasingly intense climate change has increased the frequency and intensity of extreme weather, making weather prediction critical for mitigation and adaptation. This research focuses on long-term prediction of extreme weather using the Long ShortTerm Memory (LSTM) model, as well as evaluating the influence of climate change on prediction accuracy. In this study, historical weather data is used to train and test an LSTM model combined with a RandomForestClassifier. Analysis was carried out using the Mean Squared Error (MSE) evaluation technique for 50 epochs and 8 trials at various threshold values (26, 29, 32, 35, 38, 41, 44, 47). The research results show that the LSTM model is able to predict extreme weather with an accuracy of up to 100%. Apart from that, this research also predicts daily rainfall in Bandung City through the process of data collection, preprocessing, normalization and evaluation using Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). This model produces an RMSE of 4.24 and MAE value of 2.72%, indicating quite good predictions. It is hoped that this research can make a significant contribution to the field of meteorology and can be developed further by adding parameters or other methods to improve the quality of predictions. Suggestions are given to increase the amount of data used to obtain better prediction results in the future.
Putri et al. (Mon,) studied this question.