摘要
简单介绍了SVM的理论背景,详细介绍了基于SVM的分类预测方法,给出了基于SVM的分类预测技术的性能测试结果。指出该分类预测技术可将实际问题通过非线性变换转换到高维的特征空间,在高维空间中构造线性辨别函数来实现原空间中非线性辨别函数。
This paper briefly expounds the theoretical background of SVM, intro duces in detail the methods for the class prediction based on SVM, gives the per formance measurement results of the class prediction technique based on SVM, and points out that this class prediction technique can transform the actual proble m into the high-dimensional feature space through the linear transformation, an d can construct a discriminant function in the high-dimensional space to realiz e the non-linear discriminant function in the original space.
出处
《科技情报开发与经济》
2005年第7期241-243,共3页
Sci-Tech Information Development & Economy
关键词
分类预测
向量机
机器学习
class prediction
SVM
machine learning