论文部分内容阅读
针对大数据集如何有效地进行训练的问题,基于最大向量夹角间隔分类器(maximum vector-angular margin classifier,MAMC),提出了求解最优向量d的不同方法来得到中心向量夹角间隔分类器(central vector-angular margin classifier,CAMC),进而证明了CAMC等价于最小包围球问题(minimum enclosed ball,MEB).但是鉴于MEB对参数的敏感性,又提出了正则化核向量机(regularized core vector machine,RCVM),将CAMC与RCVM结合得到中心向量夹角间隔正则化核向量机(regularized core vector machine with central vector-angular margin,CAMCVM).基于基准数据集的实验表明,CAMC具有更好的分类性能且CAMCVM可以有效快速地训练大规模数据集.
Aiming at the problem of how to train large datasets efficiently, a new method based on maximum vector-angular margin classifier (MAMC) is proposed to solve the optimal vector d to obtain the center vector angular interval classifier (CAMC), which proves that CAMC is equivalent to the minimum enclosed ball (MEB). However, given the sensitivity of the MEB to the parameters, a regularized core vector machine (regularized core vector machine (RCVM), and CAMC and RCVM are combined to obtain a center vector-angular regularized kernel vector machine (CAMCVM) .Based on dataset experiments, it is shown that CAMC has a better Categorical performance and CAMCVM can train large-scale datasets efficiently and quickly.