基于光束角思想的最大间隔学习机

Maximum margin learning machine based on beam angle

受空间几何知识和光学领域光束角的启发,提出了基于光束角思想的最大间隔学习机(BAMLM).该方法试图在模式空间中找到一个"光源"分别照射两类样本,然后根据照射区域的不同确定样本类属.分析发现,BAMLM的核化形式等价于核化中心受限最小包含球(CCMEB),通过引入核心向量机将BAMLM扩展为基于核心向量机的BAMLM(BACVM),有效地解决了大规模样本的分类问题.标准数据集和人工数据集上的实验表明了BAMLM和BACVM的有效性.

Inspired by space geometry and beam angle,a maximum margin learning machine based on beam angle（BAMLM） is proposed in this paper.The basic idea of BAMLM is to find a classified point in pattern space to separate two classes.Meanwhile,the kernelized BAMLM is equivalent to the kernelized center-constrained minimum enclosing ball（CCMEB）,and BAMLM can be extended to BACVM by introducing core vector machine（CVM） which can solve the classification for large-scale datasets.Experimental results obtained from synthetic and standard datasets show the effectiveness of the proposed algorithms.