计算机科学 ›› 2013, Vol. 40 ›› Issue (7): 206-210.
刘颖,张柏,王爱莲,桑娟,何咏梅
LIU Ying,ZHANG Bai,WANG Ai-lian,SANG Juan and HE Yong-mei
摘要: 目前,支持向量机技术(SVM)在遥感信息获取中普遍受到参数选择不准确和小样本问题的制约。针对这些问题, 提出一种新的半监督集成SVM(EPS3VM)分类模型。模型一方面利用自适应变异粒子群优化算法对SVM参数寻优以提高基分类器精度(PSVM);另一方面采用自训练算法(Self-training),充分利用大量廉价的未标记样本产生性能差异的半监督分类器个体(PS3VM),其中,在未标记样本标注过程中,引入模糊聚类算法(Gustafson-kessel)来控制错误类别的输入,最后对个体分类器采用加权集成策略,以进一步提高分类模型的泛化能力。为了测试其性能,应用该模型进行多光谱遥感影像的土地覆盖分类实验,并与PSVM、PS3VM进行对比,分类精度从PSVM的88.48%提高到96.88%,Kappa系数由0.8546提高到0.9606。结果表明,EPS3VM在克服传统SVM参数选择不准确的同时,有效地应对了小样本问题,分类性能更优。
[1] 林剑,钟迎春,彭顺喜,等.多光谱遥感图像土地利用分类区域多中心方法[J].遥感学报,2010,14(1):173-179 [2] 韩敏,林晓峰.一种基于Wedgelet变换的遥感图像分类算法[J].红外与毫米波学报,2008,27(4):280-284 [3] Knorn J,Rabe A,Radeloff V C,et al.Land cover mapping oflarge areas using chain classification of neighboring Landsat sate-llite images[J].Remote Sensing of Environment,2009,113(5):957-964 [4] Heikkinen V,Tokola T,Parkkinen J,et al.Simulated multispectral imagery for tree species classification using support vector machines[J].IEEE Transactions on Geoscience and Remote Sensing,2010,48(3):1355-1364 [5] Lardeux C,Frison P L,Tison C,et al.Support vectormachineformultifrequency SAR polarimetric data classification[J].IEEE Transactions on Geoscience and Remote Sensing,2009,47(12):4143-4152 [6] Huang Xin,Zhang Lian-pei.Road centreline extraction fromhigh-resolution imagery based on multiscale structural features and support vector machines[J].International Journal of Remote Sensing,2009,30(8):1977-1987 [7] 冯晓毅,王西博,王雷,等.基于改进JSEG算法的高分辨率遥感图像分割方法[J].计算机科学,2012,39(8):284-287 [8] 戴宏亮,戴道清.基于ETAFSVM的高光谱遥感图像自动波段选择和分类[J].计算机科学,2009,36(4):268-272 [9] LaValle S M,Branicky M S.On the relationship between classical grid search and probabilistic roadmaps[J].International Journal of Robotics Research,2002,23(8):673-692 [10] Frhlich H,Chapelle O.Feature selection for support vector machines by means of genetic algorithms[C]∥Proceedings of the 15th IEEE international conference on tools with artificial intelligence.USA,Sacramento,C A,2003 [11] Huang Cheng-lung,Dun Jian-fan.A distributed PSO-SVM hybrid system with feature selection and parameter optimization[J].Applied Soft Computing,2008,8:1381-1931 [12] Zhou Zhi-hua,Li Ming.Semi-supervised learning by disagreement[J].Knowledge and Information Systems,2010,24(3):415-439 [13] Zhou Zhi-hua.Ensemble learning[M]∥Li S Z,ed.Encyclopedia of Biometrics,Berlin:Springer,2009,0-273 [14] 邬俊,段晶,鲁明羽.基于偏袒性半监督集成的SVM主动反馈方案[J].模式识别与人工智能,2010,3(6):745-751 [15] Wu Jun,Lin Zheng-kui,Lu Ming-yu.Asymmetric Semi-Supervised Boosting for SVM Active Learning in CBIR[C]∥Procee-dings of the ACM International Conference on Image and Video Retrieval.Xi’an,China,2010 [16] Krogh A,Vedelsby J.Neural network ensembles,cross valida-tion,and active learning[J].Advances in Neural Information Processing Systems ,1995(7):231-238 [17] Liu Ying,Zhang Bai,Huang Li-hua,et al.A novel optimization parameters of support vector machines model for the land use/ cover classification[J].International Journal of Food,Agriculture & Environment,2012,10(2):132-138 [18] Maulik U,Chakraborty D.A self-trained ensemble with semisupervised SVM:An application to pixel classification of remote sensing imagery[J].Pattern Recognition,2011,44:615-623 [19] 黄金杰,李士勇,蔡云泽.一种建立粗糙数据模型的监督模糊聚类方法[J].软件学报,2005,16(6):744-753 [20] 单丹丹,杜培军,夏俊士.基于多分类器集成的“北京一号”小卫星遥感影像分类研究[J].遥感应用,2011,2:69-78 |
No related articles found! |
|