【摘要】 支持向量机是一种基于结构风险最小化原理的分类技术, 本文提出了将支持向量机分类器进行集成的分类思想。首先, 在原始样本的基础上形成子支持向量机, 得到待检样本的子预测;进而对子预测进行适当的组合, 以确定样本最终的类别预报。模拟实验结果表明, 该方法具有明显优于单一支持向量机的更高的分类准确率。更多还原

【Abstract】 The support vector machine (SVM) is a classification technique based on the structural risk minimization principle. A new classificatio method, support vector machine ensemble, is proposed in this paper. This method includes two procedures. Firstly, it gives the sub-forecast of a new sampl using sub-SVM which is obtained by Bagging or Boosting, and then, combine these sub-forecasts to decide the final class. Compared with the single suppo vector machine method, the support vector machine ensemble method has better classification accuracy.更多还原