计算机科学 ›› 2014, Vol. 41 ›› Issue (7): 283-289.doi: 10.11896/j.issn.1002-137X.2014.07.059
郭华平,袁俊红,张帆,邬长安,范明
GUO Hua-ping,YUAN Jun-hong,ZHANG Fan,Wu Chang-an and FAN Ming
摘要： 提出了一种新的基于决策树的组合分类器学习方法FL(Forest Learning)。与bagging和adaboost等传统的组合分类器学习方法不同,FL不采用抽样或加权抽样,而是直接在训练集上学习一个森林作为组合分类器。与传统组合学习方法独立地学习每个基分类器,然后把它们组合在一起的做法不同,FL学习每个基分类器时都尽可能地考虑对组合分类器的影响。首先,FL使用传统的方法构建森林的第一棵决策树；然后,逐一构建新的决策树并将其添加到森林中。在构建新的决策树时,结点的每次划分都考虑对组合分类器的影响。实验结果表明,与传统的组合分类器学习方法相比,FL在大部分数据集上都能构建出性能更好的组合分类器。
| [1] Gayar N E,Kittler J,Roli F,et al．Multiple Classifier Systems[C]∥Proceedings of 9th International Workshop on MCS．LNCS 5997,Springer 2010 [2] Sansone C,Kittler J,Roli F,et al．Multiple Classifier Systems[C]∥Proceedings of 10th International Workshop on MCS．LNCS 6713,Springer 2011 [3] Zhou Z-H,Roli F,Kittler J,et al．Multiple Classifier Systems[C]∥Proceedings of 11th International Workshop on MCS．LNCS 6713,Springer 2013 [4] Sun Y,Todorovic S,Li J,et al．Unifying the error-correcting and output-code AdaBoost within the margin framework[C]∥Raedt L D,Wrobel S,eds．Proceedings of the 22nd International Conference on Machine Learning．2005:872-879 [5] Bartlett P L,Traskin M．AdaBoost is Consistent[J]．Journal of Machine Learning Research,2007,8:2347-2368 [6] Freund Y,Schapire R F．A decision-theoretic generalization ofon-line learning and an application to boosting[J]．Journal of Computer and System Sciences,1997,55(1):119-139 [7] Dietterich T G．Ensemble methods in machine learning[C]∥Kittler J,Roli F,eds．Proceedings of the 1st International Workshop on Multiple Classifier Systems．2000:1-15 [8] Breiman L．Bagging predictors[J]．Machine Learning,1996,24(2):123-140 [9] Rodríguez J J,Kuncheva L I,Alonso C J．Rotation Forest:A New Classifier Ensemble Method[J]．IEEE Transactions on Pattern Analysis and Machine Intelligence,2006,28(10):1619-1630 [10] Breiman L．Random Forests[J]．Machine Learning,2001,45(1):5-32 [11] Zhang D,Chen S,Zhou Z-H,et al．Constraint projections for ensemble learning[C]∥Fox D,Gomes C P,eds.Proceedings of the 23rd AAAI Conference on Artificial Intelligence．Chicago,IL,2008:758-763 [12] Hecherman D．Bayesian Networks for Data Mining[J]．DataMining and Knowledge Discovery,1997,1(1):79-119 [13] Lin H-T,Li L．Support Vector Machinery for Infinite Ensemble Learning[J]．Journal of Machine Learning Research,2008,9:285-312 [14] Quinlan J R．C4.5:Programs for Machine Learning[M]．Mor-gan-kaufmann Publisher,San Mateo,CA,1993 [15] Breiman L,Friedman J H,Olshen R,et al．Classification and Regression Trees[M]． London:Chapman and & Hall,1993 [16] Ho T K．The Random Subspace Method for Constructing Deci-sion Forests[J]．IEEE Trans．Pattern Analysis and Machine Intelligence,1998,20(8):832-844 [17] Schapire R E,Freund Y,Bartlett P,et al．Boosting the Margin:A New Explanation for the Effectiveness of Voting Methods[J]．Annals of Statistics,1998,26(5):1651-1686 [18] Tan P,Steinbach M,Kumar V．数据挖掘导论[M]．范明,范宏建,等译．北京:人民邮电出版社,2008 [19] Rodríguez J J,Kuncheva L I,Alonso C J．An Experimental Study on Rotation Forest Ensembles[C]∥MCS 2007．LNCS 4472,2007:459-468 [20] Rudin C,Schapire R E,Daubechies I．Open Problem:Does AdaBoost Always Cycle?[J]．Journal of Machine Learning Research,2012,23:1-4 [21] http://nlp.zzu.edu.cn/LySpoon.asp |
| No related articles found! |
|
||
首页