文件列表:

KMBOX-0.9 (0, 2013-05-21)
KMBOX-0.9\demo (0, 2013-05-21)
KMBOX-0.9\demo\km_akcca_demo.m (4592, 2013-02-11)
KMBOX-0.9\demo\km_kcca_demo.m (3123, 2012-05-10)
KMBOX-0.9\demo\km_kcca_full_demo.m (3204, 2012-05-10)
KMBOX-0.9\demo\km_kde_demo.m (852, 2013-05-21)
KMBOX-0.9\demo\km_kpca_demo_3clusters.m (2369, 2012-02-14)
KMBOX-0.9\demo\km_kpca_demo_u.m (2667, 2012-02-14)
KMBOX-0.9\demo\km_krlst_demo.m (3525, 2012-04-29)
KMBOX-0.9\demo\km_krls_demo_mg30.m (4419, 2012-02-14)
KMBOX-0.9\demo\km_krls_demo_wiener.m (3927, 2012-02-14)
KMBOX-0.9\demo\km_krr_demo.m (1686, 2012-02-14)
KMBOX-0.9\demo\km_krr_icd_demo.m (1828, 2013-04-05)
KMBOX-0.9\demo\km_pca_demo.m (1276, 2012-02-14)
KMBOX-0.9\demo\km_template_demo.m (615, 2012-03-25)
KMBOX-0.9\km_akcca.m (15858, 2013-02-11)
KMBOX-0.9\km_aldkrls.m (3165, 2012-09-05)
KMBOX-0.9\km_fbkrls.m (4061, 2012-02-14)
KMBOX-0.9\km_kcca.m (4086, 2012-04-08)
KMBOX-0.9\km_kde.m (1426, 2013-05-21)
KMBOX-0.9\km_kernel.m (1513, 2012-05-13)
KMBOX-0.9\km_kernel_center.m (1256, 2013-02-11)
KMBOX-0.9\km_kernel_icd.m (2670, 2012-02-14)
KMBOX-0.9\km_kpca.m (1422, 2012-02-14)
KMBOX-0.9\km_kpca_icd.m (1600, 2012-02-14)
KMBOX-0.9\km_krlst.m (6193, 2012-08-21)
KMBOX-0.9\km_krr.m (1386, 2013-05-21)
KMBOX-0.9\km_norma.m (2643, 2012-02-14)
KMBOX-0.9\km_pca.m (1180, 2012-02-14)
KMBOX-0.9\km_qklms.m (3057, 2012-09-01)
KMBOX-0.9\km_silverman.m (1066, 2013-02-11)
KMBOX-0.9\km_swkrls.m (3468, 2012-06-27)
KMBOX-0.9\km_template.m (564, 2012-04-08)
KMBOX-0.9\license.txt (35146, 2010-08-31)

Kernel Methods Toolbox a MATLAB toolbox for nonlinear signal processing and machine learning Version 0.9, May 21st 2013 Author: Steven Van Vaerenbergh http://sourceforge.net/p/kmbox About ===== The Kernel Methods Toolbox (KMBOX) is a collection of MATLAB programs that implement kernel-based algorithms, with a focus on regression algorithms and online algorithms. It can be used for nonlinear signal processing and machine learning. KMBOX includes implementations of algorithms such as kernel principal component analysis (KPCA), kernel canonical correlation analysis (KCCA) and kernel recursive least-squares (KRLS). The goal of this distribution is to provide easy-to-analyze algorithm implementations, which reveal the inner mechanics of each algorithm and allow for quick modifications. The focus of these implementations is therefore on readability rather than speed or memory usage. The basis of this toolbox was a set of programs written for the Ph.D. Thesis "Kernel Methods for Nonlinear Identification, Equalization and Separation of Signals", by Steven Van Vaerenbergh, defended in 2010. Template files are provided to encourage external authors to include their own code into the toolbox. Copyright notice ================ This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 (as included and available at http://www.gnu.org/licenses). Installation ============ Extract the zip file. Add the toolbox' main folder and subfolders to the MATLAB path. Usage ===== The name of each function uses the prefix "km_" to minimize interference with other toolboxes. Usage of each function is specified in the function file itself. Most algorithms have a corresponding demonstration file in the "demo" folder that bears the suffix "_demo". These programs can be executed without setting any additional parameters. The code uses the following conventions: - For data matrices, data is stored and accessed in row format: each data point is a row in the data matrix. Extending KMBOX =============== Contributions are welcome. See contact info below. Do include at least one "demo" file for each algorithm. Citing KMBOX ============ If you use this toolbox in your research please cite @phdthesis {vanvaerenbergh2010kernel, author = {Van Vaerenbergh, Steven} title = {Kernel methods for nonlinear identification, equalization and separation of signals}, year = {2010}, school = {University of Cantabria}, month = feb, note = {Software available at \url{http://sourceforge.net/projects/kmbox/}} } Contact ======= Mail: steven (at) gtas.dicom.unican.es Web site: http://gtas.unican.es/people/steven Project web site: http://sourceforge.net/projects/kmbox/ INCLUDED ALGORITHMS =================== 1. Kernel Ridge Regression (KRR). 2. Principal Component Analysis (PCA). 3. Kernel Principal Component Analysis (KPCA), as proposed in B. Scholkopf, A. Smola and K.R. Muller, "Nonlinear component analysis as a kernel eigenvalue problem", Neural Computation, volume 10, no. 5, pages 1299- 1319, 19***. 4. Approximate Linear Dependency Kernel Recursive Least-Squares (ALD- KRLS), as proposed in Y. Engel, S. Mannor, and R. Meir. "The kernel recursive least- squares algorithm", IEEE Transactions on Signal Processing, volume 52, no. 8, pages 2275“2285, 2004. 5. Sliding-Window Kernel Recursive Least-Squares (SW-KRLS), as proposed in S. Van Vaerenbergh, J. Via, and I. Santamaria. "A sliding-window kernel RLS algorithm and its application to nonlinear channel identification", 2006 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), Toulouse, France, 2006. 6. Naive Online Regularized Risk Minimization Algorithm (NORMA), as proposed in J. Kivinen, A. Smola and C. Williamson. "Online Learning with Kernels", IEEE Transactions on Signal Processing, volume 52, no. 8, pages 2165-2176, 2004. 7. Fixed-Budget Kernel Recursive Least-Squares (FB-KRLS), as proposed in S. Van Vaerenbergh, I. Santamaria, W. Liu and J. C. Principe, "Fixed- Budget Kernel Recursive Least-Squares", 2010 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2010), Dallas, Texas, U.S.A., March 2010. 8. Incomplete Cholesky Decomposition (ICD), as proposed in Francis R. Bach and Michael I. Jordan. "Kernel Independent Component Analysis", Journal of Machine Learning Research, volume 3, pages 1-48, 2002. 9. Kernel Recursive Least-Squares Tracker (KRLS-T), as proposed in M. Lazaro-Gredilla, S. Van Vaerenbergh and I. Santamaria, "A Bayesian Approach to Tracking with Kernel Recursive Least-Squares", 2011 IEEE International Workshop on Machine Learning for Signal Processing (MLSP 2011), Beijing, ***, September, 2011. 10. Kernel Canonical Correlation Analysis (KCCA), as proposed in D. R. Hardoon, S. Szedmak and J. Shawe-Taylor, "Canonical Correlation Analysis: An Overview with Application to Learning Methods", Neural Computation, Volume 16 (12), Pages 2639--26***, 2004. 11. Quantized Kernel Least Mean Squares (QKLMS), as proposed in Chen B., Zhao S., Zhu P., Principe J.C. "Quantized Kernel Least Mean Square Algorithm," IEEE Transactions on Neural Networks and Learning Systems, vol.23, no.1, Jan. 2012, pages 22-32. 12. Alternating Kernel Canonical Correlation Analysis for blind equalization of single-input multiple-output Wiener systems, as proposed in S. Van Vaerenbergh, J. Via and I. Santamaria, "Blind Identification of SIMO Wiener Systems based on Kernel Canonical Correlation Analysis", accepted for publication in IEEE Transactions on Signal Processing, 2013. 13. Kernel density estimation (KDE). CHANGELOG ========= History of changes: v0.9 (20130521) - inclusion of KDE code and demo - minor changes v0.8 (20130211) - inclusion of AKCCA code and demo - minor changes v0.7 (20120901): - inclusion of QKLMS code - minor changes v0.6 (20120326): - inclusion of a demo for kernel canonical correlation analysis (KCCA) v0.5 (20120214): - inclusion of KRLS-T - addition of a file identifier to each file v0.4 (20110504): - inclusion of NORMA, fixed-budget KRLS, kernel PCA, incomplete Cholesky decomposition - inclusion of incomplete cholesky decomposition algorithm (km_kernel_icd). - included a listing of dependencies in function headers. - format change: dafault format for data matrices is now one data point per row (instead of one per column). - format change: one input argument less for online algorithms v0.3 (20101203): - modifications to ALD-KRLS implementation. v0.2 (20101108): - inclusion of kernel recursive least-squares algorithms (km_krls): ALD- KRLS (Approximate Linear Dependency KRLS), SW-KRLS (Sliding-Window KRLS). - correction of minor details v0.1 (20100908): - original package, includes linear PCA and kernel ridge regression algorithms.

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：