齐丙娟, 丁世飞. 基于FCM隶属度的支持向量机[J]. 微电子学与计算机, 2011, 28(10): 48-51.
引用本文: 齐丙娟, 丁世飞. 基于FCM隶属度的支持向量机[J]. 微电子学与计算机, 2011, 28(10): 48-51.
QI Bing-juan, DING Shi-fei. Support Vector Machine Based on Memberships of FCM[J]. Microelectronics & Computer, 2011, 28(10): 48-51.
Citation: QI Bing-juan, DING Shi-fei. Support Vector Machine Based on Memberships of FCM[J]. Microelectronics & Computer, 2011, 28(10): 48-51.

基于FCM隶属度的支持向量机

Support Vector Machine Based on Memberships of FCM

  • 摘要: 传统SVM在训练大规模数据集时, 训练速度慢, 时间消耗代价大.针对此问题, 提出利用FCM算法对训练样本集进行预处理, 依据样本隶属度提取出所有可能的支持向量进行SVM训练.利用原始数据集对算法进行验证, 此算法在保证SVM分类精度的同时, 大大提高了训练速度, 算法具有可行性.

     

    Abstract: As far as we know, circuit design is a process of cumbersome, complex and long cycle.This paper provides a new modeling approach based on evolutionary for circuit networks, which converts the process of network modeling into the optimization process of the determination of circuit components and component values.The new method is proposed for hybrid identification of circuit components and component values by performing global optimal search in the complex solution space and achieving the evolution of circuit networks through the way of genetic programming.

     

/

返回文章
返回