Vertical federated learning-based feature selection with non-overlapping sample utilization
作者：
Highlights：
• In this paper, we bridge this gap by proposing a novel VFL-based feature selection method—Vertical Federated Learning-based Feature Selection (VFLFS). To the best of our knowledge, this is the first deep learning-based vertical federated learning approach with feature selection.
• A strategy to make use of non-overlapping samples is also proposed to improve feature selection effectiveness.
• The proposed VFLFS approach has been evaluated extensively based on real-world datasets. The results show that VFLFS can significantly improve model performance under VFL settings compared to four state of the art baselines, especially in conditions where a large proportion of the data samples do not overlap across data owners.
摘要
•In this paper, we bridge this gap by proposing a novel VFL-based feature selection method—Vertical Federated Learning-based Feature Selection (VFLFS). To the best of our knowledge, this is the first deep learning-based vertical federated learning approach with feature selection.•A strategy to make use of non-overlapping samples is also proposed to improve feature selection effectiveness.•The proposed VFLFS approach has been evaluated extensively based on real-world datasets. The results show that VFLFS can significantly improve model performance under VFL settings compared to four state of the art baselines, especially in conditions where a large proportion of the data samples do not overlap across data owners.
论文关键词：Feature selection,Vertical federated learning,Privacy protection,Deep learning
论文评审过程：Received 17 February 2022, Revised 13 June 2022, Accepted 5 July 2022, Available online 13 July 2022, Version of Record 18 July 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.118097
