Pay Attention to Evolution: Time Series Forecasting With Deep Graph-Evolution Learning

Autor(es)

Spadon, Gabriel ; Hong, Shenda ; Brandoli, Bruno ; Matwin, Stan ; Rodrigues-Jr, Jose F. ; Sun, Jimeng

• Tipo de documento: Artigo Científico
• Fonte: IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE; v. 44, n. 9, p. 17-pg., 2022-09-01.

Resumo

Time-series forecasting is one of the most active research topics in artificial intelligence. It has the power to bring light to problems in several areas of knowledge, such as epidemiological studies, healthcare inference, and climate change analysis. Applications in real-world time series should consider two factors for achieving reliable predictions: modeling dynamic dependencies among multiple variables and adjusting the model's intrinsic hyperparameters. An open gap in the literature is that statistical and ensemble learning approaches systematically present lower predictive performance than deep learning methods. The existing applications consistently disregard the data sequence aspect entangled with multivariate data represented in more than one time series. Conversely, this work presents a novel neural network architecture for time-series forecasting that combines the power of graph evolution with deep recurrent learning on distinct data distributions, named after Recurrent Graph Evolution Neural Network (ReGENN). The idea is to infer multiple multivariate relationships between co-occurring time-series by assuming that the temporal data depends not only on inner variables and intra-temporal relationships (i.e., observations from itself) but also on outer variables and inter-temporal relationships (i.e., observations from other-selves). An extensive set of experiments was conducted comparing ReGENN with tens of ensemble methods and classical statistical ones. The results outperformed both statistical and ensemble-learning approaches, showing an improvement of 64.87 percent over the competing algorithms on the SARS-CoV-2 dataset of the renowned John Hopkins University for 188 countries simultaneously. For further validation, we tested our architecture in two other public datasets of different domains, the PhysioNet Computing in Cardiology Challenge 2012 and Brazilian Weather datasets. We also analyzed the Evolution Weights arising from the hidden layers of ReGENN to describe how the variables of the dataset interact with each other; and, as a result of looking at inter and intra-temporal relationships simultaneously, we concluded that time-series forecasting is majorly improved if paying attention to how multiple multivariate data synchronously evolve. (AU)

• Processo FAPESP: 16/17078-0 - Mineração, indexação e visualização de Big Data no contexto de sistemas de apoio à decisão clínica (MIVisBD)
• Beneficiário: Agma Juci Machado Traina
• Modalidade de apoio: Auxílio à Pesquisa - Temático
• Processo FAPESP: 20/07200-9 - Analisando dados complexos vinculados a COVID-19 para apoio à tomada de decisão e prognóstico
• Beneficiário: Agma Juci Machado Traina
• Modalidade de apoio: Auxílio à Pesquisa - Regular
• Processo FAPESP: 18/17620-5 - Medicina preventiva por meio de técnicas de deep learning aplicadas ao prognóstico de saúde
• Beneficiário: José Fernando Rodrigues Júnior
• Modalidade de apoio: Bolsas no Exterior - Pesquisa
• Processo FAPESP: 17/08376-0 - Análise e aperfeiçoamento de sistemas urbanos por meio de mapas digitais representados por redes complexas
• Beneficiário: Gabriel Spadon de Souza
• Modalidade de apoio: Bolsas no Brasil - Doutorado
• Processo FAPESP: 14/25337-0 - Desenvolvimento de algoritmos centrados em vértice para detecção de padrões em grafos em larga escala utilizando processamento paralelo assíncrono
• Beneficiário: Gabriel Perri Gimenes
• Modalidade de apoio: Bolsas no Brasil - Doutorado
• Processo FAPESP: 19/04461-9 - Prognóstico médico avançado baseado em conceitos de grafos e redes neurais artificiais
• Beneficiário: Gabriel Spadon de Souza
• Modalidade de apoio: Bolsas no Exterior - Estágio de Pesquisa - Doutorado