A stock time series forecasting approach incorporating candlestick patterns and sequence similarity
作者：
Highlights：
• Multidimensional Candlestick is encoded according to morphologies.
• Combined with future trend, the pattern of candlestick is defined.
• Sequential pattern mining approaches are extended to mine the defined pattern.
• A sequence similarity is proposed to match sequences with patterns.
• The proposed model offsets the mediocre performance of some traditional models.
摘要
•Multidimensional Candlestick is encoded according to morphologies.•Combined with future trend, the pattern of candlestick is defined.•Sequential pattern mining approaches are extended to mine the defined pattern.•A sequence similarity is proposed to match sequences with patterns.•The proposed model offsets the mediocre performance of some traditional models.
论文关键词：Stock time series forecasting,Sequential pattern mining,Candlestick pattern,Sequence similarity,Pattern matching
论文评审过程：Received 15 March 2022, Revised 21 April 2022, Accepted 12 May 2022, Available online 21 May 2022, Version of Record 7 June 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.117595
