数据流分类挖掘中的概念变化研究

摘要/Abstract

摘要： 数据流分类挖掘首先要面对概念变化问题。介绍了数据流分类中的概念变化的定义与类型,研究了概念变化的意义及应用,对目前数据流中处理概念变化的方法进行了综述。真实数据流常常含有大量的噪声,因此需要理解噪声与概念变化的区别。针对周期性数据流中概念重现现象,当“历史概念”重现时,利用特定的模型对数据流进行概念预测,可以减少模型更新的代价。

关键词: 数据流分类,概念变化,概念重现,噪声

Abstract: Data stream classification must face the concept of change.This paper introduced the definition and types of conceptual changes in the data stream classification,the meaning and application of conceptual changes,and the methods of conceptual changes in the data stream.Real data stream often contains a lot of noise,and needs to understand the difference between noise and the concept of change.To reproduce the phenomenon for periodic data stream concept,when “the concept of history” reproduces,the concept of prediction using a specific model of the data stream can reduce the model update price.

Key words: Data streams classification,Concept change,Recurrent concepts,Noise

韩法旺,刘耀宗. 数据流分类挖掘中的概念变化研究[J]. 计算机科学, 2014, 41(Z11): 347-350. https://doi.org/

HAN Fa-wang and LIU Yao-zong. Study on Concept Change in Data Streams Classification[J]. Computer Science, 2014, 41(Z11): 347-350. https://doi.org/

参考文献

[1] 王涛,李舟军,颜跃进,等.数据流挖掘分类技术综述[J].计算机研究与发展,2007,44(11):1809-1815
[2] Tsymbal A.The problem of concept drift:definitions and related work.TCD-CS-2004-15[R].Dublin,Ireland:Department of Computer Science Trinity College,Trinity College,2004
[3] Gama J.A survey on learning from data streams:current and future trends[J].Progress in Artificial Intelligence,2012,1(1):45-55
[4] 曾华军.机器学习[M].张银奎,等译.北京:机械工业出版社,2003
[5] 王永利.基于概要的数据流挖掘若干研究[D].南京:东南大学,2006
[6] 张鹏.挖掘概念漂移的数据流[D].北京:中国科学院,2009
[7] Widmer G,Kubat M.Effective learning in dynamic environ-ments by explicit context tracking[C]∥Proceedings of the Sixth European Conference on Machine Learning.Vienna,Austria,1993:69-101
[8] Schlimmer J,Granger R.Incremental learning from noisy data[J].Machine Learning,1986,1(3):317-354
[9] Li P-P,Wu X-D,Gao Y-J.A Random Decision Tree Ensemble for Mining Concept Drifts from Noisy Data Streams[J].Applied Artificial Intelligence,2010,24(7):680-710
[10] Kubat G,Gama J,Utgoff P.Special issue on incremental learning systems capable of dealing with concept drift[Z].Amsterdam,Netherlands:Intelligent Data Analysis,2004
[11] 尹志武,黄上腾.一种自适应局部概念漂移的数据流分类算法[J].计算机科学,2008,35(2):138-139
[12] 孙岳,毛国君,刘旭,等.基于多分类器的数据流中的概念漂移挖掘[J].自动化学报,2008,34(1):93-97
[13] 刘耀宗,张宏,王永利.一种自适应概念变化的数据流分类器[J].计算机研究与发展,2007,44(Suppl):63-68
[14] Klinkenberg R.Using labeled and unlabeled data to learningdrifting concepts[C]∥Proceedings of the Workshop notes of the 1 JCAI-01 Workshop on Learning from Temporal and Spatial Data.Menlo Park,CA,USA,2001:16-24
[15] Zhang P,Zhu X,Guo L.Mining data streams with labeled and unlabeled training examples[C]∥Proceedings of the ninth IEEE International Conference on Data Mining.Miami,FL,USA,2009:627-636
[16] Widmer G,Kubat M.Effective learning in dynamic environ-ments by explicit context tracking[C]∥Proceedings of the Sixth European Conference on Machine Learning.Vienna,Austria,1993:69-101
[17] Kholghi M,Keyvanpour M R.Active learning framework combining semi-supervised approach for data stream mining[J].Intelligent Computing and Information Science,2011,135:238-243
[18] Hulten G,Spencer L,Domingos P.Mining time-changing datastreams[C]∥Proceedings of the Seventh ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.San Francisco,California,USA,2001:97-106
[19] Peter V,Abranham B.Entropy-based concept drift detection[C]∥Proceedings of the 6th International Conference on Data Mining.Hong Kong,China,2006:1113-1118
[20] 张育培,柴玉梅,王黎明.基于鞅的数据流概念漂移检测方法[J].小型微型计算机系统,2013,34(8):1787-1792
[21] Verpeck J T,Meehl G A,BonyS,et al.Climate data challenges in the 21st century[J].Science,2011,331(6018):700-702
[22] 罗秀,王大玲,冯时,等.一种面向周期性概念漂移的数据流分类算法[J].计算机研究与发展,2009,6(Suppl):400-405
[23] Li P P,Wu X D,Hu X G.Mining recurring concept drifts with limited labeled streaming data[C]∥Proceedings of the 2th Asian Conference on Machine Learning.Tokyo,Japan,2010:241-252
[24] Yang Ying,Wu Xin-dong,Zhu Xing-quan.Combining proactive and reactive predictions for data streams[C]∥ Proceedings of the eleventh ACM SIGKDD International Conference on Know-ledge Discovery and Data Mining.Chicago,Illinois,USA,2005:710-715
[25] Chen S,He H,Li K,et al.MuSeRA:multiple selectively recursive approach towards nonstationary imbalanced stream data mining[C]∥Proceedings of the 2010 International Joint Confe-rence on Neural Networks.Barcelona,Spain,2010:1-8
[26] Chen S,He H.SERA:selectively recursive approach towardsnonstationary imbalanced stream data mining[C]∥Proceedings of the 2009 International Joint Conference on Neural Networks.Atlanta,Georgian,USA,2009:522-529
[27] 辛轶,郭躬德,陈黎飞,等.IKnnM-DHecoc:一种决概念漂移问题的算法[J].计算机研究与发展,2011,8(4):592-601
[28] 朱群,张玉红,胡学钢,等.一种基于双层窗口的概念漂移数据流分类算法[J].自动化学报,2011,37(9):1078-1083
[29] 杨春宇.数据流上聚类与分类算法[D].北京:清华大学,2009
[30] 周傲英,金澈清,王国仁,等.不确定性数据管理技术综述[J].计算机学报,2009,2(1):1-16

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed