基于Python的支持向量机分类算法研究.docx

经典专科、本科、硕博、研究生、期刊毕业论文 仅供参考 精心整理 仅供参考 勿用作商业用途 摘要 在大数据盛行的时代背景下，机器学习这门学科的广泛应用。并且列举运用Python语言进行数据处理的优势，将其与传统语言进行对比，充分体现了Python语言在语言简洁，效率高等方面的优势。这也是本文最后选择Python语言实现SVM算法的主要原因。 本文主体内容阐述了支持向量机算法(SVM)的基本内涵，并且用图示和数学方法形象具体讲解了SVM的基本原理。具体分析了SVM算法中线性可分数据、线性不可分数据和含有outlier点的数据集的分类方式。通过对偶问题求解法、核函数、及SMO算法等实现了对最优超平面的求解。并且完成了二分类问题向多标签分类问题的推广。 文章通过，使用手写数字数据集在Python上进行SVM模型的训练与测试，体会SVM算法如何解决实际问题。具体的表现了用Python语言实现SVM算法的优势，直观的展现了实验成果。 关键字：机器学习 Python SVM 最优超平面 核函数 Abstract In the era of the prevalence of big data, machine learning is widely used in this discipline. And enumerate the advantages of using Python language for data processing, compare it with traditional languages, and fully embody the advantages of Python language in terms of simple language and high efficiency. This is the main reason why this article chooses the Python language to implement the SVM algorithm. The main content of this paper expounds the basic connotation of Support Vector Machine (SVM), and explains the basic principle of SVM concretely with graphic and mathematical methods. The classification of linearly separable data, linearly indivisible data, and data sets with outlier points in the SVM algorithm is specifically analyzed. The solution to the optimal hyperplane is achieved through the dual problem solving method, kernel function, and SMO algorithm. And it has completed the promotion of the two-classification problem to the multi-label classification problem. Passing the end of the article, using the hand-written digital data set to train and test the SVM model in Python, to understand how the SVM algorithm solves practical problems. The specific performance of the SVM algorithm using Python language advantages, intuitive display of the experimental results. Keywords: machine learning, Python, SVM, optimal hyperplane, kernel function 第一章 绪论 1.1机器学习 随着互联网计算机技术的普及应用与发展，随处可见的数据信息数量日益庞大，数据与信息与人们的生活愈发的息息相关。数据量的不断扩大和信息获取方式的不断增多，带来了信息处理日益困难的问题。伴随着硬件性能的快速增长，人们寄希望于计算机可以帮助人类处理越来越庞大的数据。因此，机器学习在近年来得以迅速兴起与发展。 机