Extracting control variables of casting processes with NMF and rule extraction

作者：

Highlights：

• Temperate time series profiles from casting processes via Non-Negative matrix factorization

• Rule extraction from NMF results in reduced dimensional data spaces.

• Rules evaluate the physical meaning of the temperature profiles.

• Interpretable rules assist operators during casting processes.

摘要

•Temperate time series profiles from casting processes via Non-Negative matrix factorization•Rule extraction from NMF results in reduced dimensional data spaces.•Rules evaluate the physical meaning of the temperature profiles.•Interpretable rules assist operators during casting processes.

论文关键词：Non-negative matrix factorization (NMF),Rule extraction,Interpretability,Industrial process monitoring

论文评审过程：Received 24 January 2021, Revised 2 April 2021, Accepted 22 April 2021, Available online 28 April 2021, Version of Record 10 May 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115118