Statistical mechanics of support vector networks
Statistical mechanics of support vector networks
复制标题
DOI：
10.1103/physrevlett.82.2975
复制
发表时间：
1999-04-05
影响因子：
8.6
通讯作者：
Sompolinsky, H
中科院分区：
文献类型：
--
作者：
Dietrich, R;Opper, M;Sompolinsky, H
使用统计物理学方法，我们研究了支持向量机（SVM）的泛化性能，支持向量机最近被引入作为神经网络的通用替代品。对于非线性分类规则，当示例数量太少而无法正确估计非线性部分的系数时，泛化误差会达到饱和状态。当使用简单规则进行训练时，我们发现 SVM 的过拟合程度很弱。当输入的分布在特征空间中存在间隙时，SVM 的性能会大大增强。 [S0031-9007(99)08788-8]。
Using methods of statistical physics, we investigate the generalization performance of support vector machines (SVMs), which have been recently introduced as a general alternative to neural networks. For nonlinear classification rules, the generalization error saturates on a plateau when the number of examples is too small to properly estimate the coefficients of the nonlinear part. When trained on simple rules, we find that SVMs overfit only weakly. The performance of SVMs is strongly enhanced when the distribution of the inputs has a gap in feature space. [S0031-9007(99)08788-8].