| 基于SVM的网络文本信息自动分类 |
| |
| 引用本文： | 刘清,陈炼,吕静. 基于SVM的网络文本信息自动分类[J]. 现代计算机, 2007, 0(10): 14-16,57 |
| |
| 作者姓名： | 刘清 陈炼 吕静 |
| |
| 作者单位： | [1]南昌大学计算中心,南昌330031 [2]江西财经大学,南昌330031 |
| |
| 基金项目： | 江西省教育厅科技项目计划 |
| |
| 摘 要： | 介绍基于SVM的网络文本信息自动分类算法,该算法在训练阶段将一个大型数据集分成许多不相交的子集,按批次对各个训练子集中的样本进行训练而得到多个分类器,利用误差纠错输出编码优化分类器,从而减少较深层次训练需要学习的文档.
|
| 关 键 词： | 支持向量机 自动分类 多层分类 误差纠正输出编码 |
| 修稿时间： | 2007-07-172007-09-05 |
SVM-based Hypertext Information Automatic Categorization |
| |
| LIU Qing,CHEN Lian,LV Jing. SVM-based Hypertext Information Automatic Categorization[J]. Modem Computer, 2007, 0(10): 14-16,57 |
| |
| Authors: | LIU Qing CHEN Lian LV Jing |
| |
| Affiliation: | 1. Computer Center, Nanchan University,Nanchan 330031; 2. Jiangxi University of Finance and Economics, Nanchan 330031 |
| |
| Abstract: | Introduces an algorithm of hypertext information categorization based on SVM, the algorithm divides a large data set into many non-intersecting subsets during training period, in which the samples are trained according to the batch and then many classifications are constructed. The classifications are optimized by error correcting output codes(ECOC), which reduces the amount of documents to be studied in the deep level training phase. |
| |
| Keywords: | Support Vector Machine Categorization Multi-level Classification Error Correcting Output Codes |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
