信息向量机是一种基于贝叶斯理论的稀疏高斯过程方法,其模型训练速度快、内存耗费小、稀疏性强,具有良好的预测性能。本文从高斯过程回归模型出发,提出了一种基于信息向量机的高光谱影像分类方法,针对高斯过程分类中的非高斯噪声模型,采用假定概率滤波算法将分类问题转化为回归问题,通过最大化边缘似然函数进行模型训练,选择活动子集中的信息向量数量达到了稀疏的目的。通过ROSIS影像试验,表明了基于信息向量机的高光谱影像分类方法的优势。
Informative vector machine is a method of sparse Gaussian process based on Bayesian theory, which has high speed in model training, small consuming in memory, strong effective in sparseness and good forecasting performance. In this paper, the Gaussian process regression model is introduced firstly, and then a hyperspectral imagery classification method based on informative vector machine is brought forward. Secondly, to solve the problem of non-Gaussian noise model in the Gaussian process classification, the classification problem is transformed into a regression problem by using the assume density filtering algorithm, after which model is trained by maximizing the marginal likelihood function. Finally, the number of informative vector is chosen in active subset to achieve the purpose of sparse. According to the experimental results of ROSIS images, the advantages of hyperspectral imagery classification method based on informative vector machine are validated.
[1] YANG Guopeng, YU Xuchu, FENG Wufa, et al. The Development and Application of Hyperspectral RS Technology[J]. Bulletin of Surveying and Mapping, 2008(10):1-4.(杨国鹏, 余旭初, 冯伍法, 等. 高光谱遥感技术的发展与应用现状[J]. 测绘通报, 2008(10):1-4.)
[2] TAN Xiong, YU Xuchu, ZHANG Pengqiang, et al. A Classification Algorithm for Hyperspectral Images Based on Fuzzy Mixed Pixel Decomposition[J]. Journal of Geomatics Science and Technology, 2013, 30(3):279-283.(谭熊, 余旭初, 张鹏强, 等. 一种基于模糊混合像元分解的高光谱影像分类方法[J]. 测绘科学技术学报, 2013, 30(3):279-283.)
[3] YU Xuchu, FENG Wufa, YANG Guopeng, et al. Analysis and Application for Hyerspectral Imagery[M]. Beijing:Science Press, 2013.(余旭初, 冯伍法, 杨国鹏, 等. 高光谱影像分析与应用[M]. 北京:科学出版社, 2013.)
[4] TAN Xiong. Research on Classification Techniques for Hyperspectral Imagery Based on Combined Spectral and Spatial Features[D]. Zhengzhou:Information Engineering University, 2014.(谭熊. 联合光谱和空间特征的高光谱影像分类技术研究[D]. 郑州:信息工程大学, 2014.)
[5] SUN Weiwei. Theory and Methods of Dimensionality Reduction Using Manifold Learning for Hyperspectral Imagery[J]. Acta Geodaetica et Cartographica Sinica, 2014, 43(4):439.(孙伟伟. 基于流形学习的高光谱影像降维理论与方法研究[J]. 测绘学报, 2014, 43(4):439.)
[6] SHI Qian, DU Bo, ZHANG Liangpei. A Dimensionality Reduction Method for Hyperspectral Imagery Based on Local Discriminative Tangent Space Alignment[J]. Acta Geodaetica et Cartographica Sinica, 2012, 41(3):417-420.(石茜, 杜博, 张良培. 一种基于局部判别正切空间排列的高光谱遥感影像降维方法[J]. 测绘学报, 2012, 41(3):417-420.)
[7] TAN Kun, DU Peijun. Wavelet Support Vector Machines Based on Reproducing Kernel Hilbert Space for Hyperspectral Remote Sensing Image Classification[J]. Acta Geodaetica et Cartographica Sinica, 2011, 40(2):142-147.(谭琨, 杜培军. 基于再生核Hilbert空间的小波核函数支持向量机的高光谱遥感影像分类[J]. 测绘学报, 2011, 40(2):142-147.)
[8] ZHANG Lei, SHAO Zhenfeng, ZHOU Xiran, et al. Semi-supervised Collaborative Classification for Hyperspectral Remote Sensing Image with Combination of Cluster Feature and SVM[J]. Acta Geodaetica et Cartographica Sinica, 2014, 43(8):855-861.(张磊, 邵振峰, 周熙然, 等. 聚类特征和SVM组合的高光谱影像半监督协同分类[J]. 测绘学报, 2014, 43(8):855-861.)
[9] YANG Guopeng, YU Xuchu, ZHOU Xin, et al. Research on Relevance Vector Machine for Hyperspectral Imagery Classification[J]. Acta Geodaetica et Cartographica Sinica, 2010, 39(6):572-577.(杨国鹏, 余旭初, 周欣, 等. 基于相关向量机的高光谱影像分类研究[J]. 测绘学报, 2010, 39(6):572-577.)
[10] YAO Futian, QIAN Yuntao. Gaussian Process and Its Applications in Hyperspectral Image Classification[J]. CAAI Transactions on Intelligent Systems, 2011, 6(5):396-404.(姚伏天, 钱沄涛. 高斯过程及其在高光谱图像分类中的应用[J]. 智能系统学报, 2011, 6(5):396-404.)
[11] CSATÓ L. Gaussian Processes-iterative Sparse Approximations[D]. Aston:Aston University, 2002.
[12] LAWRENCE N D, SEEGER M, HERBRICH R. The Informative Vector Machine:A Practical Probabilistic Alternative to the Support Vector Machine[R]. Sheffield, UK:Technical Report, Department of Computer Science, 2005.
[13] LAWRENCE N D, SEEGER M, HERBRICH R. Fast Sparse Gaussian Process Methods:The Informative Vector Machine[M]//Advances in Neural Information Processing Systems.[S.l.]:MIT Press, 2003:625-632.
[14] LAWRENCE N D, PLATT J C. Learning to Learn with the Informative Vector Machine[C]//Proceedings of the 21st International Conference in Machine Learning. San Francisco:[s.n.], 2004:512-519.
[15] LIU Jianwei, XU Xiang, LUO Xionglin. Face Recognition Based on Orthogonal Locality Preserving Projection and Informative Vector Machine[J]. Computer Engineering, 2010, 36(7):200-202.(刘建伟, 徐翔, 罗雄麟. 基于OLPP和信息向量机的人脸识别[J]. 计算机工程, 2010, 36(7):200-202.)
[16] XU Xiang, LIU Jianwei, LUO Xionglin. Research on Nerval Activity Classification and Decoding Based on Informative Vector Machine[J]. Computer Engineering, 2010, 36(7):198-199, 202.(徐翔, 刘建伟, 罗雄麟. 基于信息向量机的神经活动分类和译码研究[J]. 计算机工程, 2010, 36(7):198-199, 202.)
[17] GIBBS M N, MACKAY D J C. Variational Gaussian Process Classifiers[J]. IEEE Transactions on Neural Networks, 2002, 11(6):1458-1464.
[18] NEAL R M. Regression and Classification Using Gaussian Process Priors[J]. Bayesian Statistics, 1998, 6(10):475-501.
[19] WILUAMS C K I, BARBER D. Bayesian Classification with Gaussian Processes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1998, 20(12):1342-1351.
[20] MINKA T P. A Family of Algorithms for Approximate Bayesian Inference[D]. Cambridge:Massachusetts Institute of Technology, 2001:36-48.
[21] MØLLER M F. A Scaled Conjugate Gradient Algorithm for Fast Supervised Learning[J]. Neural Networks, 1993, 6(4):525-533.
[22] LI Hang. Statistical Learning Method[M]. Beijing:Tsinghua University Press, 2012.(李航. 统计学习方法[M]. 北京:清华大学出版社, 2012.)
