SAX-ARM: Deviant event pattern discovery from multivariate time series using symbolic aggregate approximation and association rule mining

作者：

Highlights：

• SAX-ARM algorithm is proposed for discovering rules from multivariate time series.

• Inverse normal transformation (INT) is adopted for normalizing time series.

• Symbolic aggregate approximation (SAX) is applied to discretize time series.

• Association rule mining (ARM) discovers frequent patterns among deviant events.

• A die-casting manufacturing dataset is used to illustrate the propose method.

摘要

•SAX-ARM algorithm is proposed for discovering rules from multivariate time series.•Inverse normal transformation (INT) is adopted for normalizing time series.•Symbolic aggregate approximation (SAX) is applied to discretize time series.•Association rule mining (ARM) discovers frequent patterns among deviant events.•A die-casting manufacturing dataset is used to illustrate the propose method.

论文关键词：Multivariate time series,Event pattern discovery,Inverse normal transformation (INT),Symbolic aggregate approximation (SAX),Association rule mining (ARM)

论文评审过程：Received 26 February 2019, Revised 8 August 2019, Accepted 12 September 2019, Available online 13 September 2019, Version of Record 26 September 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.112950