Minimum class variance multiple kernel learning

作者：

Highlights：

• We propose an improved MKL method, which exploits the ellipsoidal structure of the data.

• We develop two optimization strategies to handle the optimization model of the proposed method.

• We discuss a variant that can automatically tune the regularization parameter.

• We conduct the comprehensive experiments to evaluate the effectiveness of the proposed methods.

摘要

•We propose an improved MKL method, which exploits the ellipsoidal structure of the data.•We develop two optimization strategies to handle the optimization model of the proposed method.•We discuss a variant that can automatically tune the regularization parameter.•We conduct the comprehensive experiments to evaluate the effectiveness of the proposed methods.

论文关键词：Kernel methods,Multiple kernel learning,Support vector machine,Structure information

论文评审过程：Received 18 January 2020, Revised 23 June 2020, Accepted 1 September 2020, Available online 16 September 2020, Version of Record 18 September 2020.

论文官网地址：https://doi.org/10.1016/j.knosys.2020.106469