文件列表:

SemiSupFeat_toolbox (0, 2013-05-23)
SemiSupFeat_toolbox\COPYING.txt (1432, 2012-12-11)
SemiSupFeat_toolbox\contents.m~ (2514, 2013-05-23)
SemiSupFeat_toolbox\contents.m (2464, 2013-05-23)
SemiSupFeat_toolbox\tools (0, 2013-05-23)
SemiSupFeat_toolbox\tools\generate_toydata.m (5350, 2012-12-12)
SemiSupFeat_toolbox\tools\binarize.m (196, 2012-02-08)
SemiSupFeat_toolbox\tools\tile.m (3586, 2010-12-02)
SemiSupFeat_toolbox\tools\bagkernelImGMM.m (1617, 2012-12-12)
SemiSupFeat_toolbox\tools\closerClusterGMM.m (746, 2012-12-13)
SemiSupFeat_toolbox\tools\kernelmatrix.m (1183, 2012-02-08)
SemiSupFeat_toolbox\tools\plotKernelFeatures.m (1382, 2012-12-13)
SemiSupFeat_toolbox\tools\scalestd.m (255, 2012-02-08)
SemiSupFeat_toolbox\tools\plotFeatures.m (1328, 2012-12-12)
SemiSupFeat_toolbox\tools\estimateSigma.m (1920, 2012-02-08)
SemiSupFeat_toolbox\tools\gen_eig.m (1032, 2012-03-24)
SemiSupFeat_toolbox\tools\kernelcentering.m (3420, 2011-09-19)
SemiSupFeat_toolbox\tools\bagkernelgmm.m (2502, 2012-12-12)
SemiSupFeat_toolbox\tools\figure_projections.m (530, 2012-12-12)
SemiSupFeat_toolbox\methods (0, 2013-05-23)
SemiSupFeat_toolbox\methods\Assesspls.m (1667, 2013-05-23)
SemiSupFeat_toolbox\methods\dualpls.m (2408, 2012-02-08)
SemiSupFeat_toolbox\methods\Assesspca.m (1840, 2013-05-23)
SemiSupFeat_toolbox\methods\assessment.m (4775, 2012-02-08)
SemiSupFeat_toolbox\methods\pca.m (535, 2013-05-23)
SemiSupFeat_toolbox\methods\Assesskpls.m (5020, 2013-05-23)
SemiSupFeat_toolbox\methods\isp_rkopls_phase1.m (4038, 2012-03-24)
SemiSupFeat_toolbox\methods\plsSB.m (732, 2012-12-28)
SemiSupFeat_toolbox\methods\Assesskopls.m (6033, 2013-05-23)
SemiSupFeat_toolbox\methods\isp_rkopls_phase3.m (1375, 2012-02-24)
SemiSupFeat_toolbox\DemoSemiFeat.m (3888, 2013-05-23)
SemiSupFeat_toolbox\.directory (40, 2012-12-13)

SEMIFEAT_TOOLBOX AUTHORS: E. Izquierdo-Verdiguier, L. Gomez-Chova, G. Camps-Valls FILES: % Assesskopls - Computes the overall accuracy OA and kappa index. The estimated labels are obtained by projected data onto principal components of KOPLS method. % Assesskpls - Computes the overall accuracy OA and kappa index. The estimated labels are obtained by projected data onto principal components of KPLS method. % assessment - Computes quantitative measures of accuracy of classifiers and regression methods % Assesspca - Computes the overall accuracy OA and kappa index. The estimated labels are obtained by projected data onto principal components of PCA method. % Assesspls - Computes the overall accuracy OA and kappa index. The estimated labels are obtained by projected data onto principal components of PLS method. % bagkernelgmm - Builds a GMM cluster kernel given input labeled and unlabeled data % bagkernelImGMM - Builds different kinds of supervised and unsupervised kernels % binarize - Binarize of a vector % closerClusterGMM - Finds the closer clusters for generating test kernels % contents - General purpose commands. % DemoSemiFeat - Demo in order to comparing the richness of extraction with differents linear and nonlinear feature extraction methods: PCA, PLS, OPLS, CCA, MNF, their corresponding nonlinear methods and KECA. % dualpls - Performs dual (kernel) PLS discrimination [www.kernel-methods.net,J. Shawe-Taylor and N. Cristianini, Kernel Methods for Pattern Analysis, Cambridge University Press, 2004.] % estimateSigma - Estimates the sigma parameter (RBF kernel lengthscale) from available data. % figure_projections - Plot projected data % gen_eig - Extracts generalized eigenvalues for problem A * U = B * U * Landa. % generate_toydata - Generate original data. % isp_rkopls_phase1 - Implements the OPLS method for FE projecting the input data either in original space or in feature space (KOPLS). Phase 1 : Sequential computation. % isp_rkopls_phase3 - Extracts Features from data. % kernelcentering - function Kc=kernelcentering(K,Ktrain) % kernelmatrix - function K = kernelmatrix(ker,X,X2,sigma) % pca - Compute the principal components of PCA method. % plotFeatures - Plot projections % plotKernelFeatures - Plot projections of kernel-based methods % pls - Compute the principal components of PLS method. % scalestd - Standardizes each variable in the dataset independently to have zero mean and unit variance % stratify - Stratified v-fold: Split the v-folds such that they have the same number of class labels each Code for the papers: E. Izquierdo-Verdiguier, L. Gomez-Chova, L. Bruzzone, G. Camps-Valls "Semisupervised nonlinear feature extration for image classification" IEEE International Geoscience and Remote Sensing Symposium, IGARSS’2012 Munich, Germany, 2012 E. Izquierdo-Verdiguier, J. Arenas-Garcia, S. Munoz-Romero L. Gomez-Chova, G. Camps-Valls "Semisupervised Kernel Orthonormalized Partial Least Squares" IEEE International Workshop on Machine Learning for Signal Processing, MLSP’12 Santander, Spain, 2012 REFERENCES * BOOKS * I. T. Jollife, Principal Component Analysis, Springer, 1***6. J. Shawe-Taylor and N. Cristianini, Kernel Methods for Pattern Analysis, Cambridge University Press, 2004. C. M. Bishop, Pattern Recognition and Machine Learning (Information Science and Statistics), Springer-Verlag New York, Inc., 2006. O. Chapelle, B. Scholkopf, and A. Zien, Semi-Supervised Learning, MIT Press, Cambridge, 1st edition, 2006. G. Camps-Valls and L. Bruzzone, Kernel Methods for Remote Sensing Data Analysis, John Wiley and Sons, 2009. * PAPERS * M. A. Kramer, “Nonlinear principal component analysis using autoassociative neural networks,” AIChE Journal, vol. 37, no. 2, pp. 233–243, 1991. R. Rosipal and N. Kramer, “Overview and recent advances in partial least squares,” in Subspace, Latent Structure and Feature Selection. 2006, vol. 3940 of LNCS, pp. 34–51, Springer. S. T. Roweis and L. K. Saul, “Nonlinear dimensionality reduction by locally linear embedding,” Science, vol. 290, no. 5500,pp. 2323–2326, December 2000. R. Rosipal and L.J. Trejo, “Kernel partial least squares regression in reproducing kernel Hilbert space,” J. Mach. Learn. Res., vol. 2, pp. 97–123, March 2002. J. Arenas-Garcia and G. Camps-Valls, “Efficient kernel orthonormalized PLS for remote sensing applications,” IEEE Trans. Geosc. Rem. Sens., vol. 46, pp. 2872 –2881, Oct 2008. D. Tuia and G. Camps-Valls, “Urban image classification with semisupervised multiscale cluster kernels,” IEEE JSTARS, vol. 4, pp. 65–74, Mar 2011. D. Tuia, F. Ratle, A. Pozdnoukhov, and G. Camps-Valls, “Multisource composite kernels for urban-image classification,” IEEE Geosc. Rem. Sens. Lett., vol. 7, pp. 88–92, 2010. M. L. Braun, J. Buhmann, and K. R. Muller, “On relevant dimensions in kernel feature spaces,” Journal of Machine Learning Research, vol. 9, pp. 1875–1908, Aug 2008. G. Camps-Valls, J. Mooij, and B. Scholkopf, “Remote sensing feature selection by kernel dependence measures,” IEEE Geosc. Rem. Sens. Lett., vol. 7, no. 3, pp. 587–591, 2010. J. Arenas-Garcia, K. B. Petersen, and L. K. Hansen, “Sparse kernel orthonormalized PLS for feature extraction in large datasets,” in NIPS19. 2007, MIT Press. D. Tuia and G. Camps-Valls, “Semisupervised remote sensing image classification with cluster kernels,” IEEE Geosc. Rem. Sens. Lett., vol. 6, no. 2, pp. 224–228, Apr 2009. O. Chapelle, J. Weston, and B. Scholkopf, “Cluster Kernels for Semi-Supervised Learning,” in NIPS 2002, Becker, Ed., Cambridge, MA, USA, 2003, vol. 15, pp. 585–592, MIT Press. J. Arenas-Garcia and G. Camps-Valls, “Efficient kernel orthonormalized PLS for remote sensing applications,” IEEE Trans. Geosc. Rem. Sens., vol. 46, pp. 2872 –2881, Oct 2008. Y. Gu, Y. Liu, and Y. Zhang, “A selective KPCA algorithm based on high-order statistics for anomaly detection in hyperspectral imagery,” IEEE Geosci. Remote Sens. Letters, vol. 5, no. 1, pp. 43 –47, Jan 2008. B. C. Kuo, C. H. Li, and J. M. Yang, “Kernel nonparametric weighted feature extraction for hyperspectral image classification,” Geosci. Remote Sens., IEEE Transactions on, vol. 47, no. 4, pp. 1139 –1155, apr. 2009. H. Wold, “Estimation of principal components and related models by iterative least squares.,” Multivariate Analysis, pp. 391–420, 1966.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：