期刊文献+

基于自适应集成分类器的数据流概念漂移算法 被引量:6

下载PDF
导出
摘要 数据流具有连续、实时、有序及无限等特点,使用传统的数据挖掘技术来处理数据流的分类面临着严重的挑战,很难处理数据流中的概念漂移问题。文章结合现有的决策树分类挖掘算法,提出了自适应集成分类器方法,构建了数据流概念漂移的自适应集成分类模型,通过不断更新训练样例的权重与属性类别,将训练样例从现有的数据集中分离出来,并被确定为新类别属性的训练样例,以达到对数据流中概念漂移现象的有效检测,仿真结果也证明该方法的适应性和可靠性。
出处 《统计与决策》 CSSCI 北大核心 2016年第7期13-17,共5页 Statistics & Decision
基金 教育部博士点基金资助项目(20123718120004) 全国统计科研计划项目(2012LY183) 山东省软科学项目(2014RKB01506)
  • 相关文献

参考文献13

  • 1Turne V D. Cost-sensitive Classification: Empirical Evaluation of a Hybrid Genetic Decision Tree Induction Algorithm [J]. Journal of Arti- ficial Intelligence Research, 1995, (2).
  • 2Elwell R , Polikar R. Incremental Learning of Concept Drift in Non-stationary Environments[C]. 1EEE Transactions on Neural Net- works, 2011.
  • 3Farid D M, Rahman C M. Assigning Weights to Training Instances In- creases Classification Accuracy[J]. International Journal of Data Min- ing & Knowledge Management Process, 2013, (3).
  • 4Su L, Liu H Y , Song Z H. A New Classification Algorithm for Data Stream[J].International Journal of Modern Education and Computer Science, 2011, (4).
  • 5Masud M, Kevin W. Facing the Reality of Data Stream Classification: Coping with Scarcity of Labeled Data [J].Knowledge and Information Systems, 2012, (33).
  • 6Loh W Y, Shih X. Split Selection Methods for Classification Tree [J]. Statistica Sininca, 1997, (7).
  • 7张杰,孙曰瑶.基于AdaBoost组合算法的衍生金融工具风险预测[J].统计与决策,2012,28(7):41-44. 被引量:12
  • 8Farid D M , Rahman C M. Assigning Weights to Training Instances Increases Classification Accuracy [J]. International Journal of Data Mining Knowledge Management Process, 2013, (3).
  • 9Read J , Pfahringer B. Scalable and Efficient Multi-label Classifica- tion for Evolving Data Streams [J].Machine Learning, 2012, (88).
  • 10Biswas A. Farid D M. A new Decision Tree Learning Approach for Novel Class Detection in Concept Drifting Data Stream Classifica- tion [J].The Journal of Computer Science and Engineering, 2012, (14),.

二级参考文献18

  • 1Freund Y,Iyer R,Schapirer R E,et al.An Efficient Boosting Algo-rithm for Combining Preferences[J].Journal of Machine Learning Re-search,2003,(4).
  • 2Li,H.,Sun,J.Majority Voting Combination of Multiple Case-based Reasoning for Financial Distress Prediction[J].Expert Systems with Applications,2009,(36).
  • 3Beaver,W.Financial Ratios as Predictors of Failure[J].Journal of Ac-counting Research,1966,(4).
  • 4Sun,J.,Li,H.Financial Distress Prediction Based on Serial Combina-tion of Multiple Classifiers[J].Expert Systems with Applications,2009,(36).
  • 5Domingos P, Hulten G. Mining high-speed data streams[C]. Proc of ACM Sigkdd Int Conf Knowledge Discovery in Databases. Boston: ACM Press, 2000: 71-80.
  • 6Hulten G, Spencer L, Domingos E Mining time-changing data streams[C]. Proc of ACM Sigkdd Int Conf Knowledge Discovery in Databases. San Francisco: ACM Press, 2001: 97-106.
  • 7Wang H, Fan W, Yu P, et al. Mining concept drifting data streams using ensemble classifiers[C]. The 9th ACM lnt Conf on Knowledge Discovery and Data Mining. Washington: ACM Press, 2003: 226-235.
  • 8Tom Mitchell. Machine learning[M]. McGraw Hill, 1997: 123_12~6.
  • 9Zico Kolter J, Marcus A Maloof. Dynamic weighted majority: An ensemble method for drifting concepts[J]. J of Machine Learning Research, 2007, 8(8): 2755-2790.
  • 10Li C Q, L!ng T W, Hu M. Efficient processing of updates in dynamic XML data[C]. Proc of the 22nd Int Conf on Data Engineering. Washington DC: IEEE Computer Society, 2006: 13-22.

共引文献22

同被引文献47

引证文献6

二级引证文献32

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部