Alternating multiconlitron: A novel framework for piecewise linear classification
作者：
Highlights：
• Alternating multiconlitron is a novel framework for piecewise linear classification.
• It is based on the concept of maximal convexly separable subset (MCSS).
• It can be constructed by support alternating multiconlitron algorithm (SAMA).
• It generally has a much simpler structure than multiconlitron.
• Experiments verify effectiveness and competitiveness of the proposed method.
摘要
Highlights•Alternating multiconlitron is a novel framework for piecewise linear classification.•It is based on the concept of maximal convexly separable subset (MCSS).•It can be constructed by support alternating multiconlitron algorithm (SAMA).•It generally has a much simpler structure than multiconlitron.•Experiments verify effectiveness and competitiveness of the proposed method.
论文关键词：Multiconlitron,Alternating multiconlitron,Maximal convexly separable subset,Support alternating multiconlitron algorithm,Piecewise linear classifier
论文评审过程：Received 27 March 2014, Revised 17 August 2014, Accepted 6 September 2014, Available online 16 September 2014.
论文官网地址：https://doi.org/10.1016/j.patcog.2014.09.003
