一种有效的大规模数据的分类方法

An Efficient Target Recognition Method for Large Scale Data

本文提出了一种基于自组织特征映射神经网络 (SOM)和支撑矢量机 (SVM)相结合的复杂模式的大规模数据的分类方法 .该方法首先利用自组织特征映射神经网络对待识目标进行聚类 ,然后应用支撑矢量机方法对其进行分类识别 .通过对复杂异或 (XOR)分类问题 ,以及实际的Iris和Appendicitis数据分类问题等的分类实验 ,且与仅用支撑矢量机的分类方法比较 ,结果表明 ,本文提出的方法对复杂模式的大规模数据的分类识别问题具有较好的效果 ,且训练时间大幅度减小 .

An efficient target recognition method for large scale data is proposed in this paper,which is based on self-organizing map (SOM) neural network and support vector machines (SVMs).The target data set is divided into clusters by SOM first.Then,the support vector machines are applied to classify targets.The new method is used to classify the complex XOR problem,Iris and Appendicitis data,and the experimental results show that the new method can obtain better recognition results for the complex pattern classification of large scale data,and the trainning time is shorter than that by using the support vector machine method only.