Directed Acyclic Graph networks to characterize phase space evolution with application to musical composition and industrial maintenance

作者：

Highlights：

• A new transition network based on causality is proposed to process time series.

• The new method is used for system identification and characterization.

• Entropy properties of the new network’s detect system behavioural transitions.

• The potential of the method is demonstrated on music and real-world industrial data.

摘要

•A new transition network based on causality is proposed to process time series.•The new method is used for system identification and characterization.•Entropy properties of the new network’s detect system behavioural transitions.•The potential of the method is demonstrated on music and real-world industrial data.

论文关键词：Directed acyclic graph (DAG),Permutation entropy,Behavioural change detection,Complex networks

论文评审过程：Received 1 December 2021, Revised 8 August 2022, Accepted 13 August 2022, Available online 19 August 2022, Version of Record 27 August 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118586