COVID-19 CASE FORECASTING USING NEURAL NETWORK ALGORITHMS (PREVISÃO DE CASOS DE COVID-19 UTILIZANDO ALGORITMOS DE REDES NEURAIS)

Resumo

 https://doi.org/10.5281/zenodo.15671164

This paper explores the prediction of new COVID-19 cases using recurrent neural networks (RNNs) such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU). Data from the São Paulo State Data Analysis System (SEADE) was utilized. Additionally, a scalable cloud architecture using AWS was employed. The data underwent pre-processing techniques like scaling and moving average. Following processing, the models were evaluated using MAE (Mean Absolute Error) and R² metrics. Results indicate that the GRU model exhibited superior performance, likely attributed to its ability to efficiently handle low-dimensional data.