作者: Li Zhang , Xingning Lu , Bangjun Wang , Shuping He
DOI: 10.1109/IJCNN.2015.7280325
关键词:
摘要: Similarity learning ranges over an extensive field in machine and pattern recognition. This paper deals with similarity based on multiple support vector data description (SVDD). It is well known that SVDD was proposed for one-class or two-class unbalanced problems. Thus, we propose a (MSVDD) algorithm apply it to multi-class A model trained by similar pairwise samples the same class instead of all ones. In addition, dissimilar are not considered MSVDD. Experimental results validate MSVDD promising learning.