上海大学2013 ~ 2014 学年 冬 季学期研究生课程考试小论文格式课程名称:模式识别方法课程编号: 09SB59004 论文题目 : 支持向量机原理及应用研究生姓名 : 章云元学 号: 13721278 论文评语 : 成 绩: 任课教师 : 李昕评阅日期 : 支持向量机原理及应用学号: 13721278 姓名:章云元日期: 2014.03.07 摘要: 支持向量机是从统计学发展而来的一种新型的机器学习方法,在解决小样本、非线性和高维的机器学习问题中表现出了许多特有的优势,但是,支持向量机方法中也存在着一些亟待解决的问题,主要包括:如何用支持向量机更有效的解决多类分类问题,如何解决支持向量机二次规划过程中存在的瓶颈问题、如何确定核函数以及最优的核参数以保证算法的有效性等。本文详细介绍系统的阐述了统计学习理论、支持向量机理论以及支持向量机的主要研究热点,包括求解支持向量机问题、多类分类问题、参数优化问题、核函数的选择问题等。关键词: 机器学习;统计学习理论;SVM;VC 维;The principle and application of Support Vector Machine ABSTRACT : SVM(Support Vector Machine) is a novel method of machine learning evolving from Statistics. SVM presents many own advantages in solving machine learning problems such as small samples, nonlinearity and high dimension. However, SVM methods exist some problems need to be resolved, mainly including how to deal with multi-classification effectively, how to solve the bottle-neck problem appearing in quadratic programming process, and how to decide kernel function and optimistical kernel parameters to guarantee effectivity of the algorithm.This paper has introduced in detail the structure, evolvement history, and kinds of classification of machine learning, and demonstrated system SLT(Statistical Learning Theory), SVM and research hotspots of SVM, including seeking SVM problems, multi-classification, parameters optimization, kernel function selection and so on. Keywords: Machine learning, SLT, SVM, VC dimension 1. 引言1.1 支持向量机研究背景及意义随着...