摘要
支持向量机(SVM)是解决小样本学习问题的有力工具,其关键是如何得到判别样本类别的最优超平面。受约束条件的限制,最优超平面的求解比较繁琐。遗传算法具有全局搜索最优解的特点,是求最优值问题的非常有效的方法。由此,利用遗传算法得到了一个直接求最优超平面近似解的方法,该方法不同于传统的二次规划方法。
<Abstrcat> Support Vector Machine is a powerful instrument which can solve the small sample study problem.The key problem of the support vector machine (SVM) is how to obtain an optimal hyperplane that is the critical boundary for distinguishing the sample which belongs to one kind or another.Genetic algorithm is a method to search the optimal solution globally,with which a direct method for searching an optimal hyperplane is given.It is different from the quadratic programing method.
出处
《上海工程技术大学学报》
CAS
2005年第1期21-23,共3页
Journal of Shanghai University of Engineering Science
关键词
支持向量机
最优超平面
遗传算法
适应度函数
support vector machines
optimal hyperplane
genetic algorithm
fitness function
