Quantitative Association Rules Applied to Climatological Time Series Forecasting

This work presents the discovering of association rules based on evolutionary techniques in order to obtain relationships among correlated time series. For this purpose, a genetic algorithm has been proposed to determine the intervals that form the rules without discretizing the attributes and allow...

ver descrição completa

Detalhes bibliográficos
Autores: Martínez Ballesteros, María del Mar, Martínez Álvarez, Francisco, Troncoso Lora, Alicia, Riquelme Santos, José Cristóbal
Formato: capítulo de livro
Estado:Versión publicada
Fecha de publicación:2009
País:España
Recursos:Universidad de Sevilla (US)
Repositorio:idUS. Depósito de Investigación de la Universidad de Sevilla
OAI Identifier:oai:idus.us.es:11441/40508
Acesso em linha:http://hdl.handle.net/11441/40508
https://doi.org/10.1007/978-3-642-04394-9_35
Access Level:acceso abierto
Palavra-chave:Time series
Forecasting
Quantitative association rules
Descrição
Resumo:This work presents the discovering of association rules based on evolutionary techniques in order to obtain relationships among correlated time series. For this purpose, a genetic algorithm has been proposed to determine the intervals that form the rules without discretizing the attributes and allowing the overlapping of the regions covered by the rules. In addition, the algorithm has been tested on real-world climatological time series such as temperature, wind and ozone and results are reported and compared to that of the well-known Apriori algorithm.