Feature extraction of auto insurance size of loss data using functional principal component analysis

作者：

Highlights：

• Dimension reduction and feature exaction approach via functional PCA are used.

• Patterns and variability of extracted features of the size-of-loss are investigated.

• The distribution fitting for irregularly grouped data is studied.

• The proposed method is capable of capturing the tail behaviour of the distribution.

摘要

•Dimension reduction and feature exaction approach via functional PCA are used.•Patterns and variability of extracted features of the size-of-loss are investigated.•The distribution fitting for irregularly grouped data is studied.•The proposed method is capable of capturing the tail behaviour of the distribution.

论文关键词：Explainable data analysis,Data visualization,Functional principal component analysis,Loss distribution,Business analytics,Spline methods

论文评审过程：Received 21 July 2020, Revised 6 January 2022, Accepted 25 February 2022, Available online 7 March 2022, Version of Record 17 March 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116780