Diarrhoea outpatient visits prediction based on time series decomposition and multi-local predictor fusion

作者：

Highlights：

• A EEMDAN–GRNN algorithm is proposed for diarrhoea outpatient visits prediction.

• Predictor based on decomposition-and-ensemble principle superior to single predictor.

• GRNN is a good candidate as predictor for diarrhoea outpatient visits prediction.

• EEMDAN is superior to EMD and Wavelet for diarrhoea outpatient visits time series.

摘要

•A EEMDAN–GRNN algorithm is proposed for diarrhoea outpatient visits prediction.•Predictor based on decomposition-and-ensemble principle superior to single predictor.•GRNN is a good candidate as predictor for diarrhoea outpatient visits prediction.•EEMDAN is superior to EMD and Wavelet for diarrhoea outpatient visits time series.

论文关键词：Diarrhoea outpatient visits,Prediction model,Time series decomposition,Ensemble empirical mode decomposition,Generalized regression neural networks,Multi-predictor fusion

论文评审过程：Received 24 December 2014, Revised 2 August 2015, Accepted 4 August 2015, Available online 25 August 2015, Version of Record 11 September 2015.

论文官网地址：https://doi.org/10.1016/j.knosys.2015.08.013