摘要
支持向量机中参数设置对训练支持向量机分类的精确度有不可忽视的影响。支持向量机参数的选取可看作参数的组合优化。免疫算法是一种有效的随机全局优化技术,它具有不易陷入局部最优解、解精度高、收敛速度快等优点。该文利用人工免疫算法进行支持向量机模型选择。该算法主要包括克隆选择、高频变异、受体编辑等操作。试验证明,该算法能够有效提高支持向量机分类的正确性。
The parameters setting for Support Vector Machine（SVM） in a training process impacts on the classification accuracy. The selection problem of SVM parameters is considered as a compound optimization problem. Immune algorithm is an efficient random global optimization technique. It has nice performances such as avoiding local optimum, high precision solution, and quick convergence. This paper proposes an immune algorithm applied to model selection of SVM. This algorithm includes clonal selection, hyper-mutation and receptor editing. Experimental results indicate that this method significantly improves the classification accuracy of SVM.
出处
《计算机工程》
CAS
CSCD
北大核心
2008年第15期223-225,共3页
Computer Engineering
关键词
支持向量机
模型选择
免疫算法
Support Vector Machine（SVM）
model selection
immune algorithm
