Period-aware local modelling and data selection for time series prediction
作者：
Highlights：
• The introduced algorithm selects useful data for improved training of local models.
• A hybrid usefulness-related distance is proposed for training data selection.
• Data usefulness is evaluated by taking into account periodicity of time series.
• Autocorrelation function and Renyi entropy is used to reduce number of parameters.
• The proposed method offers lower prediction error than the state-of-the-art local and global models.
摘要
•The introduced algorithm selects useful data for improved training of local models.•A hybrid usefulness-related distance is proposed for training data selection.•Data usefulness is evaluated by taking into account periodicity of time series.•Autocorrelation function and Renyi entropy is used to reduce number of parameters.•The proposed method offers lower prediction error than the state-of-the-art local and global models.
论文关键词：Local models,Time series prediction,Data reduction,Segmentation,k-nearest neighbours,Soft computing
论文评审过程：Received 25 May 2015, Revised 2 March 2016, Accepted 18 April 2016, Available online 20 April 2016, Version of Record 30 April 2016.
论文官网地址：https://doi.org/10.1016/j.eswa.2016.04.024
