计算机科学 ›› 2017, Vol. 44 ›› Issue (Z11): 381-384.doi: 10.11896/j.issn.1002-137X.2017.11A.080
周国华,巢海鲸,申燕萍
ZHOU Guo-hua, CHAO Hai-jing and SHEN Yan-ping
摘要: 迁移学习方法是一种新的机器学习框架,它将源领域数据通过学习迁移到相似的目标领域中,减弱了对已标记数据的依赖。但迁移学习方法中一个重大问题是使用目标领域数据与源领域数据得到的分类器很可能比仅利用目标领域数据得到的分类器的效果更差,从而造成一种“负迁移”现象。针对此问题,提出一种基于目标领域已标记数据知识的安全控制机制,并通过结合近年出现的一种迁移学习分类器(TL-SVM)提出了一种安全迁移支持向量机(SATL-SVM),从理论上解决了TL-SVM的负迁移问题,在人工数据集和真实数据集上的实验结果表明了所提方法的有效性。
[1] PAN S J,YANG Q.A survey on transfer learning[J].IEEE Transactions on Knowledge & Data Engineering,2010,22(10):1345-1359. [2] 庄福振,罗平,何清等.迁移学习研究进展[J].软件学报,2015,6(1):26-39. [3] WEISS K,TAGHI M K,WANG D D.A survey of transferlearning[J].Journey of Big Data,2016,3(9):1-40. [4] LU J,BEHBOOD V,HAO P,et al.Transfer learning using com-putational intelligence:a survey[J].Knowledge-Based Systems,2015,80(5):14-23. [5] BIONDI G O,PRATI R C.Setting parameters for support vector aachines using transfer learning[J].Journal of Intelligent & Robotic Systems,2015,80(12):295-311. [6] YING L,LIU B.Application of transfer learning in task recommendation system[J].Procedia Engineering,2017,174(2):518-523. [7] OPBROEK V,IKRAM A.Transfer learning improves super-vised image segmentation across imaging protocols[J].IEEE Transactions on Medical Imaging,2015,34(5):1018-1030. [8] YANG C J,DENG Z H,CHOI K S,et al.Takagi-Sugeno-Kang transfer learning fuzzy logic system for the adaptive recognition of epileptic electroencephalogram signals[J].IEEE Transactions on Fuzzy Systems,2016,24(5):1079-1094. [9] CHENG B,LIU M X,ZHANG D Q,et al.Domain transfer learning for MCI conversion prediction[J].IEEE Transactions on Biomedical Engineering,2015,62(7):1805-1817. [10] MEI S Y.SVM ensemble based transfer learning for large-scale membrane proteins discrimination[J].Journal of TheoreticalBiology,2014,340(1):105-110. [11] UGENT J D,BURM M,KINDERMANS P J.Transfer learning of gaits on a quadrupedal robot[J].Adaptive Behavior,2015,23(2):69-82. [12] GAO J,FAN W,JIANG J,et al.Knowledge transfer via multiple model local structure mapping[C]∥ACM the 14th SIGKDD International Conference on Knowledge Discovery and Data Mining.New York,USA:ACM,2008:283-291. [13] 洪佳明,印鉴,黄云,等.一种基于领域相似性的迁移学习算法[J].计算机研究与发展,2011,48(10):1823-1830. [14] 许敏,王士同,顾鑫.TL-SVM:一种迁移学习新算法[J].控制与决策,2014,29(1):141-146. [15] VAPNIK V.Statistical Learning Theory[M].John Wiley and Sons,1998. [16] ARGYRIOU A,MICCHELLI C A,PONTIL M.When is there a representer theorem? vector versus matrix regularizers [J].Journal of Machine Learning Research,2009,10(12):2507-2529. [17] 邓乃杨,田英杰.数据挖掘的新方法——支持向量机[M].北京:科学出版杜,2004. [18] XIANG E W,CAO B,HU D H,et al.Bridging domains using world wide knowledge for transfer learning[J].IEEE Transactions on Knowledge and Data Engineering,2010,22(6):770-783. [19] BRUZZONE L,MARCONCINI M.Domain adaptation prob-lems:A DASVM classification technique and a circular validation strategy[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2010,32(5):770-787. [20] CHANG C C,LIN C J.LIBSVM:a library for support vector machines[J/OL].http://www.csie.ntu.edu.tw/~cjlin/li-bsvm,2001. |
No related articles found! |
|