Selective multiple kernel learning for classification with ensemble strategy

作者：

Highlights：

• We obtain a competitive result with MKL, meanwhile owning sparsity.

• We propose a new kernel evaluation method with quantified result.

• We save the memory to optimize MKL and extend the scale of problem.

• We accelerate MKL optimization by using Lp-norm(p≥2).

• A fast SMKL with L∞-norm is proposed, without MKL optimization.

摘要

Highlights•We obtain a competitive result with MKL, meanwhile owning sparsity.•We propose a new kernel evaluation method with quantified result.•We save the memory to optimize MKL and extend the scale of problem.•We accelerate MKL optimization by using Lp-norm(p≥2).•A fast SMKL with L∞-norm is proposed, without MKL optimization.

论文关键词：Ensemble learning,Kernel evaluation,Multiple kernel learning,Selective multiple kernel learning,Fast selective multiple kernel learning

论文评审过程：Received 23 December 2011, Revised 31 January 2013, Accepted 7 April 2013, Available online 24 April 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.04.003