Publicação

Forecasting natural gas prices using a hybrid deep learning model and news

Ver documento

Detalhes bibliográficos
Resumo:The transition to cleaner energy in the European Union prioritizes natural gas, yet the Russo-Ukrainian War caused unpredictable price fluctuations. Our study aimed to enhance predictive models by exploring GDELT data, analyzing pre- and post-war performance, and comparing deep learning models (RNN, LSTM, GRUNN). Incorporating crude oil and average tone data significantly improved predictions. Geopolitical factors necessitate further research to ensure energy security and economic development. Employing CRISP-DM methodology, we established a systematic approach to address these challenges. Our study contributes valuable insights to enhance predictions and adapt models to complex energy markets.
Autores principais:Filho, René Alexandre Porto da Franca Rocha
Assunto:Natural gas Price prediction GDELT News sentiment Hybrid Deep learning Gás natural Previsão preço Sentimentos de notícias Híbrido
Ano:2023
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:ISCTE
Idioma:inglês
Origem:Repositório ISCTE
Descrição
Resumo:The transition to cleaner energy in the European Union prioritizes natural gas, yet the Russo-Ukrainian War caused unpredictable price fluctuations. Our study aimed to enhance predictive models by exploring GDELT data, analyzing pre- and post-war performance, and comparing deep learning models (RNN, LSTM, GRUNN). Incorporating crude oil and average tone data significantly improved predictions. Geopolitical factors necessitate further research to ensure energy security and economic development. Employing CRISP-DM methodology, we established a systematic approach to address these challenges. Our study contributes valuable insights to enhance predictions and adapt models to complex energy markets.