何顺兵数学模型论文.doc

PAGE PAGE 1 处理非线性分类和回归问题的一种新方法(I) 支持向量机方法简介 陈永义 俞小鼎 高学浩 冯汉中” (中国气象局培训中心，北京100081) (成都气象中心，成都610071) 摘 要 简要介绍了近年来倍受瞩目的一种处理高度非线性分类、回归等问题的计算机学习的新方法——支持向量机(SVM)方法；分析了这一方法的特点及其在数值预报产品释用及气象研究业务中的应用前景。SVM是一种有坚实理论基础的新颖的小样本学习方法。它基本上不涉及概率测度及大数定律等，因此不同于现有的统计方法。从本质上看，它避开了从 归纳到演绎的传统过程，实现了高效的从训练样本到预报样本的“转导推理”(transductive inference)，大大简化了通常的分类和回归等问题。SVM的最终决策函数只由少数的支持向量所确定，计算的复杂性取决于支持向量的数目，而不是样本空间的维数，这在某种意义上 避免了“维数灾”。 关键词：机器学习 支持向量机(SVM) 模式识别 回归估计天气预报 ABSTRACT A brief introduction to an increasingly popular machine learning technique，SVM (support vector machine)is presented for solving nonlinear classification and regress on problems·Properties of SVM are discussed together with potentials of applying SVM to numerical weather forecast． SVM is a novel learning method that has solid theoretical basis and rebut reson1v same amount of sample． It does not rely on probability measures and Law to Large Numbers．hence is different from many other statistical methods·In essence，SVM m televised the traditional inference process from induction to deflection- Instead， it employs transductive inference from training sample to predicting sample， which greatly simplifies classification and regression problems．The decision function of SVM is only determined by a few support vectors．The complexity of computation relates to the humbler of support Vectors rather than the dimension of the sample space．Thus，to some degree， SVM avoids the curse 0f dimensionally”． KEY WORDS：Machine learning ,Support Vector machines(SVM), Pattern recognition, Regression estimation, Weather forecast  目录 引言……………………………………………………………..3 1 类和回归问题的提法…………………………………………5 2 优化分线性超平面和支持向量 ……………………………………………….5 3 性支持向量机………………………………………………..7 3.1线性可分离的情况………………………………………….7 3.2线性不可分的情况………………………………………….8 3.3线性多类分类问题………………………………………….8 4 非线性支持向量机…………………………………………..9 4.1Mercer核和Mercer定理…………………………………….9 4.2特征映射和特征空间……………………………………….10 4.3非线性支持向量机………………………………………….10 5 SVM回归方法简介…………………………………………….11 6 VM方法的特点及应用展望……