基于W^1 2再生核支持向量机的模式分类研究

支持向量机是基于统计学习理论的模式分类器。它通过结构风险最小化准则和核函数方法，较好地解决了模式分类器复杂性和推广性之间的矛盾，引起了大家对模式识别领域的极大关注。近年来，支持向量机在手写体识别、人脸识别、文本分类等领域取得了很大的成功。文章将一种新的核函数用于虹膜识别，并与传统的多项式核函数、高斯核函数进行了比较。初步结果显示了该核函数的应用潜力。

Pattern Recognition Research of Support Vector Machine Based on Reproducing Kernel of W^1 2

Support vector machine is a kind of pattern classifier based on statistical learning theory. By structuring risk minimization principle and kernel function method, it solves the contradictions between the complicacy and generalization of pattern classifier preferably, which arouses people concentrating on the domain of pattern identification. Recent years, support vector machine has obtained great success in the fields of handwriting identification, human face identification, text classification and so on. This paper applies a new kind of kernel function to iris identification, and compares with traditional polynomial kernel function and Gauss kernel function. The principal result shows the potential application of this kernel function.