Multi-dimensional classification via kNN feature augmentation
作者：
Highlights：
• A first attempt towards feature manipulation in input space for multi-dimensional classification (MDC) is proposed.
• The original feature space of MDC examples is enriched with augmented features based on kNN techniques.
• Extensive experimental studies over 15 benchmark data sets have been conducted for performance evaluation.
摘要
•A first attempt towards feature manipulation in input space for multi-dimensional classification (MDC) is proposed.•The original feature space of MDC examples is enriched with augmented features based on kNN techniques.•Extensive experimental studies over 15 benchmark data sets have been conducted for performance evaluation.
论文关键词：Machine learning,Multi-dimensional classification,Feature augmentation,Class dependencies
论文评审过程：Received 30 January 2020, Revised 28 March 2020, Accepted 4 May 2020, Available online 16 May 2020, Version of Record 23 May 2020.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107423
