A Comparative Study of RNN and DNN for Climate Prediction

In this article, the authors conducted a comparative study of two artificial neural network models, recurrent neural network (RNN) with long short-term memory (LSTM) and deep neural network (DNN), for the prediction of daily variations of temperature, precipitation, and humidity data in a specific geographic area in Morocco. The study aimed to assess the effectiveness of the two models in climate prediction. The results indicated that both models performed well in making daily predictions, but the LSTM RNN showed superior performance in making weekly predictions. The study contributes valuable insights into the application of deep learning models for climate prediction, highlighting the potential of RNN with LSTM for capturing long-term dependencies in noisy datasets like climate variables.