摘要
针对标签均值半监督支持向量机在图像分类中随机选取无标记样本会导致分类正确率不高,以及算法的稳定性较低的问题,提出了基于聚类标签均值的半监督支持向量机算法。该算法修改了原算法对于无标记样本的惩罚项,对选取的无标记样本聚类,使用聚类标签均值替换标签均值。实验结果表明,使用聚类标签均值训练的分类器大大减少了背景与目标的错分情况,提高了分类的正确率以及算法的稳定性,适合用于图像分类。
Semi-supervised support vector machine(S3 VM)based on label mean can lead to low classification accuracy and unstable results due to random selection of unlabeled samples.In order to deal with the problems,we propose a semi-supervised support vector machine based on clustering label mean.This method modifies the penalty terms of the original algorithm for unlabeled samples,clusters unlabeled samples and replaces label mean with clustering label mean.Experimental results indicate that the proposed method greatly reduces the misclassification of background and objectives,improves the stability and classification accuracy of the algorithm,and it is suitable for image classification.
引文
[1] Zhu Xiao-jin.Semi-supervised learning literature survey[D].Madison:University of Wisconsin,2016.
[2] Bennett K,Demiriz A.Semi-supervised support vector machines[C]∥Proc of Advances in Neural Information Processing Systems,1998:368-374.
[3] Belkin M,Niyogi P,Sindhwani V.Manifold regularization:A geometric framework for learning from labeled and unlabeled examples[J].The Journal of Machine Learning Research,2006(7):2399-2434.
[4] Gómez-Chova L,Camps-Valls G,Munoz-Mari J,et al.Semisupervised image classification with Laplacian support vector machines[J].IEEE Geoscience and Remote Sensing Letters,2008,5(3):336-340.
[5] Chen Yi-song,Wang Guo-ping,Dong Shi-hai.Learning with progressive transductive support vector machine[J].Pattern Recognition Letters,2003,24(12):1845-1855.
[6] Li Y F,Kwok J T,Zhou Z H.Semi-supervised learning using label mean[C]∥Proc of the 26th Annual International Conference on Machine Learning,2009:633-640.
[7] Zeng S,Huang R,Kang Z,et al.Image segmentation using spectral clustering of Gaussian mixture models[J].Neurocomputing,2014,144:346-356.
[8] Du P J,Tan K,Xing X S.Wavelet SVM in reproducing kernel Hilbert space for hyperspectral remote sensing image classification[J].Optics Communications,2010,283(24):4978-4984.
[9] Li M H,Liao C M.Sensitivity analysis for a Lagrange dual problem to a vector optimization problem[J].Optimization Letters,2013,7(8):1837-1846.
[10] Wu Q,Zhou D X.SVM soft margin classifiers:Linear programming versus quadratic programming[J].Neural Computation,2005,17(5):1160-1187.
[11] Wu Zi-li.KKT conditions for weak compact convex sets,theorems of the alternative,and optimality conditions[J].Journal of Functional Analysis,2014,266(2):693-712.
[12] Wang Shuo-chen,Wang Xi-li,Ma Jun-liang.Semi-supervised support vector machine for image classification based on mean shift[J].Journal of Computer Applications,2014,34(8):2399-2403.(in Chinese)
[13] Borenstein E.Weizmann horse database[EB/OL].[2013-08-20].http:∥www.msri.org/people/members/eranb/.
[14] LIBSVM—A library for support vector machines[EB/OL].[2018-07-15].http:∥www.csie.ntu.edu.tw/~cjlin/libsvm.
[12]王朔琛,汪西莉,马君亮.基于均值漂移的半监督支持向量机图像分类[J].计算机应用,2014,34(8):2399-2403.