Abstract: This paper presents a novel classifier called Total Margin v-Fuzzy Support Vector Machine(TM-v-FSVM) based on the idea of FSVM and Total Margin(TM). Although it can be seen as a modified class of the classic FSVM, TM-v-FSVM has better theoretical classification performance than FSVM. The proposed method solves not only the overfitting problem resulted from outliers with the approaches of fuzzification of the penalty and total margin algorithm, but also the imbalanced datasets by using different cost algorithm, thus obtaining a lower generalization error bound. Exprimental results obtained with UCI real datasets show that the algorithm is more stable and superior than other related diagrams.

Key words: Total Margin(TM), generalization, support vector, Fuzzy Support Vector Machine(FSVM), linearity, nonlinearity

摘要： 为获得更好的分类性能，对传统模糊支持向量机(FSVM)进行扩展，提出一种总间隔v-模糊支持向量机(TM-v-FSVM)。通过使用差异成本及引入总间隔和模糊隶属度，同时解决不平衡训练样本问题和传统软间隔分类机的过拟合问题，从而提升学习机的泛化能力。采用UCI实际数据集进行模式分类实验，结果表明TM-v-FSVM具有稳定的分类性能。

关键词: 总间隔, 泛化, 支持向量, 模糊支持向量机, 线性, 非线性

CLC Number: 

• TP181