Geographical discrimination of propolis using dynamic time warping kernel principal components analysis
作者：
Highlights：
• PCA cannot handle GC–MS data well since it does not account for mis-aligned peaks.
• Gaussian dynamic time warping (DTW) kernel is proposed to align GC–MS data.
• The proposed method reveals better clusters of propolis from 8 regions in the US.
• The proposed DTW-KPCA of GC–MS data is more effective than classic PCA and Kernel PCA.
摘要
•PCA cannot handle GC–MS data well since it does not account for mis-aligned peaks.•Gaussian dynamic time warping (DTW) kernel is proposed to align GC–MS data.•The proposed method reveals better clusters of propolis from 8 regions in the US.•The proposed DTW-KPCA of GC–MS data is more effective than classic PCA and Kernel PCA.
论文关键词：Clustering,Machine learning,GC–MS,Time-series alignment,Kernel method
论文评审过程：Received 13 January 2021, Revised 2 June 2021, Accepted 17 September 2021, Available online 24 September 2021, Version of Record 28 September 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.115938
