核选择是支撑向量机(Support Vector Machine,SVM)研究中的核心问题之一。提出了一种基于数据分布特征的SVM核函数及其参数选择的方法。首先分析了确定数据分布特征的重要性,然后给出了判断数据呈高斯分布的方法,并探讨了SVM核函数及其参数选择与数据分布的相关性。数值实验说明了本文提出的方法的可行性与有效性。
The kernel selection is one of the key problems for support vector machine (SVM). In this paper, a new way to select the kernel parameter, is presented. It is based on the Gauss distribution. The paper analyses the importance of determining the characteristics of data distribution and presents an approach to determine Gauss distribution. And then on the basis of determining Gauss distribution ,this paper discusses how to select the kernel parameter. The simulation experiments demonstrate the feasibility and the effectiveness of the presented approach.