A novel time series clustering method with fine-tuned support vector regression for customer behavior analysis

作者：

Highlights：

• A two-stage customer behavior analysis methodology is proposed.

• The methodology performs customer segmentation using historical transactions.

• Hybrid SVRGOA is proposed to forecast customers’ behavior accurately.

• The empirical results show that the SVRGOA outperforms other standard techniques.

摘要

•A two-stage customer behavior analysis methodology is proposed.•The methodology performs customer segmentation using historical transactions.•Hybrid SVRGOA is proposed to forecast customers’ behavior accurately.•The empirical results show that the SVRGOA outperforms other standard techniques.

论文关键词：Customer relationship management (CRM),Customer behavior,Feature-based time series clustering,Time series forecasting,Support vector regression

论文评审过程：Received 31 January 2022, Revised 21 March 2022, Accepted 10 May 2022, Available online 17 May 2022, Version of Record 22 May 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117584