Hierarchical classification method based on selective learning of slacked hierarchy for activity recognition systems
作者：
Highlights：
• The problem of multi-class activity classification is addressed in this article.
• The proposed method is based on error-correcting output coding (ECOC) method.
• Two independent base classifiers are combined for multi-class activity recognition.
• A multistage maximum voting process is used to combine the results of classifiers.
• The proposed method is verified to be applicable for real-time sensor readings.
摘要
•The problem of multi-class activity classification is addressed in this article.•The proposed method is based on error-correcting output coding (ECOC) method.•Two independent base classifiers are combined for multi-class activity recognition.•A multistage maximum voting process is used to combine the results of classifiers.•The proposed method is verified to be applicable for real-time sensor readings.
论文关键词：Multi-class classification,Hierarchical classification,Error-correcting output coding,Activity recognition,Slacked Hierarchy
论文评审过程：Received 18 July 2016, Revised 31 May 2017, Accepted 27 June 2017, Available online 28 June 2017, Version of Record 7 July 2017.
论文官网地址：https://doi.org/10.1016/j.eswa.2017.06.040
