基于神经网络集成分类器预处理的支持向量机分类算法

A Support Vector Machine Algorithm Based on Pretreatment of Neural Network Ensemble

由于支持向量机分类算法对于小样本数据效果不是很好,本文提出了一种基于神经网络集成分类器预处理的支持向量机分类算法。该算法首先通过神经网络集成分类器扩充样本集,然后利用支持向量机分类算法对新样本集合进行学习。由于神经网络集成分类器可以较好地扩充样本集合,所以可以有效地提高支持向量机分类算法训练的精度。在UCI标准数据集上的实验表明,基于神经网络集成分类器预处理的支持向量机分类算法较传统的支持向量机算法具有更高的精度。

As the support vector machine classification algorithm is not very good on small sample data set,we proposed a support vector machine algorithm based on pretreatment of neural network ensemble.This algorithm first expands the data set by neural network ensemble classifier.Then use the support vector machine classification algorithm to study on the new sample set.As Neural network ensemble classifier can expand the sample set effectively,so the classification accuracy of support vector machines can be improved effectively.Experiments on the UCI standard data set showed that,compared with the classical support vector machine algorithm,the proposed algorithm has higher accuracy.