中文说明：支持向量机(Support Vector Machine)是Cortes和Vapnik于1995年首先提出的，它在解决小样本、非线性及高维模式识别中表现出许多特有的优势，并能够推广应用到函数拟合等其他机器学习问题中。支持向量机方法是建立在统计学习理论的VC 维理论和结构风险最小原理基础上的，根据有限的样本信息在模型的复杂性（即对特定训练样本的学习精度，Accuracy）和学习能力（即无错误地识别任意样本的能力）之间寻求最佳折衷，以期获得最好的推广能力（或称泛化能力）。matlab的SVM源代码，包括分类、拟合和回归。
English Description：
SVM (Support Vector Machine) is made by the Cortes and Vapnik in 1995 was the first, in solving nonlinear and multidimensional pattern recognition of small samples, shows many unique advantages and fitting can be applied to functions and other machine learning problems. Support vector machine method is established in statistics learning theory of VC dimension theory and structure risk minimum principle based Shang of, according to limited of sample information in model of complexity (that on specific training sample of learning precision, Accuracy) and learning capacity (that no errors to recognition arbitrary sample of capacity) Zhijian sought best folding compromise, to obtained best of promotion capacity (or said generalization capacity). SVM MATLAB source code, including classification, fitting and regression.