| 摘要: |
| 针对传统神经网络存在网络结构难于确定、过学习以及局部极小等问题，研究了基于支持向量机（SVM）的模式识别问题。通过对棋盘这种典型非线性二值问题的分类研究，分析了支持向量机的分类与泛化能力。支持向量机在分类和泛化能力方面远远优于传统神经网络。最后将支持向量机用于对两类飞机目标的分类识别，通过多组蒙特卡罗试验，获得了较好的识别结果。支持向量机在目标识别中有巨大潜力和广阔前景。 |
| 关键词: 模式识别 支持向量机 径向基函数 泛化能力 目标识别 |
| DOI：10.3969/j.issn.1001-893X. |
| 投稿时间：2005-12-14修订日期：2006-04-04 |
| 基金项目:国家自然科学基金 |
|
| Application of Support Vector Machine(SVM) in Pattern Recognition |
|
| () |
| Abstract: |
| Aiming at the problems such as difficult determination of net structure,over fitting and local minimization of traditional neural networks,the support vector machine(SVM) applied to pattern recognition is studied.By investigating the chessboard classification,which is typical of nonlinear two-value problem,the generalization ability of SVM is analyzed. SVM is more powerful than traditional neural network in the aspect of classification and generalization.Finally two kinds of airplanes are recognized based on SVM,with many Monte-Carlo experiments good classification results are achieved.SVM has huge potentials and good prospect in the area of target recognition. |
| Key words: pattern recognition,support vector machine(SVM),range profiles,generalization ability,target recognition |
《电讯技术》编辑部 