基于增量式学习的数据流实时分类模型

被引:4
作者
孙娜
郭延锋
机构
[1] 辽宁工业大学电子与信息工程学院
关键词
增量式学习; 支持向量机; 网络异常检测; 概念漂移; 多分类器模型;
D O I
10.16208/j.issn1000-7024.2012.11.073
中图分类号
TP181 [自动推理、机器学习];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
传统数据挖掘方法,主要针对静态数据进行挖掘,而对数据流挖掘往往失效。为了解决数据流的数据挖掘问题,提出一种通过改变传统支持向量机增量式学习方法,利用轮转式结构将多分类器按照数据流时间顺序进行组合,并且通过对分类器的优化,可以提高模型对数据流分类的准确率并减少训练时间消耗。实验结果表明,该模型在保证学习精度和推广能力的同时,提高了训练速度,适合于数据流在线分类和在线学的问题。
引用
收藏
页码:4225 / 4229
页数:5
相关论文
共 11 条
[1]  
Minimax probability machine with geneticfeature optimized for intrusion detection. Chen Z-G,Wang S. InformationTechnology Journal . 2008
[2]   快速支持向量机增量学习算法 [J].
赵耀红 ;
王快妮 ;
钟萍 ;
王来生 .
计算机工程与设计, 2010, 31 (01) :161-163+171
[3]  
Facial expression recognitionusing improved support vector machine by modifying kernels. Liejun W,Xizhong Q,Taiyi Z. Information Technology Journal . 2009
[4]  
Research on credit scoring model with SVM for network management. Feng L,Yao Y,Jin B. Journal of Computational Information System . 2010
[5]  
Evaluation of Credit Risk Based on Firm Performance European. Maria Psillaki. Journal of operational Research . 2010
[6]  
On the sparseness of 1-norm support vector machines. Zhang L,Zhou W. Neural Networks . 2010
[7]  
Least squares support vector machine on Morlet wavelet kernel function and itsapplication to nonlinear system identification. Wu Fangfang,,Zhao Yinliang. Information Technology Journal . 2006
[8]  
An incremental learning algorithm for Lagrangian support vector machines. Duan Hua,Shao Xiao jian,Hou Wei zhen et al. Pattern Recognition . 2009
[9]   An incremental learning algorithm for Lagrangian support vector machines [J].
Duan, Hua ;
Shao, Xiaojian ;
Hou, Weizhen ;
He, Guoping ;
Zeng, Qingtian .
PATTERN RECOGNITION LETTERS, 2009, 30 (15) :1384-1391
[10]  
Detection of LSB matchingsteganography using neighborhood node degree characteristics. Xia B,Sun X,Xiang L,et al. Information Technology Journal . 2011