| 引用本文： |
-
吴建生,金龙.神经网络的统计学习理论基础[J].广西科学院学报,2005,(2):102-105,109. [点击复制]
- Wu Jiansheng,Jing Long.The Theory Elements of Neural Network Statistical Learning[J].Journal of Guangxi Academy of Sciences,2005,(2):102-105,109. [点击复制]
|
|
| 摘要: |
| 介绍神经网络的统计学习过程和理论,讨论基于经验风险最小化的学习理论对神经网络推广性能的影响,分析基于结构风险最小化的支持向量机.认为神经网络因其出色的高度非线性映射能力、自组织和适应能力、记忆联想能力,使得神经网络成为机器学习的重要研究领域. |
| 关键词: 神经网络 学习过程 经验风险 结构风险 支持向量机 |
| DOI： |
| 投稿时间：2004-12-20修订日期：2005-03-03 |
| 基金项目:广西自然科学基金(0339025)资助项目。 |
|
| The Theory Elements of Neural Network Statistical Learning |
|
Wu Jiansheng1, Jing Long2
|
| (1.Dept. of Math. & Comp., Liuzhou Teacher Coll., Liuzhou, Guangxi, 545004, China;2.Guangxi Research Inst. of Meteorological Disasters Mitigation, Nanning, Guangxi, 530022, China) |
| Abstract: |
| The statistical learning process and theory of the neural network are introduced.The influence of generation ability based on the empirical risk minimization and the support vector machines based on the structural risk minimization are discussed.The neural network becomes a research hotspot in machine learning because of its outstanding nonlinear mapping,self-organized,parallelity, adaptation. |
| Key words: neural network learning process empirical risk structural risk support vector machines |
