wkl

所属分类:matlab编程
开发工具:matlab
文件大小:110KB
下载次数:27
上传日期:2014-07-24 21:41:27
上 传 者bobo2017
说明:  小波核方法工具箱,来自法国University de Rouen的Alain Rakotomamonjy 教授,有对应论文Wavelet Kernel Learning。
(This package is a set of Matlab scripts that implements the algorithms described in the submitted paper: "Wavelet Kernel Learning". )

文件列表:
toyBench (0, 2010-09-01)
Comparespeed (0, 2010-09-01)
toyBench\setpath.m (1143, 2010-09-01)
toyBench\WklEeg2.m (1264, 2010-04-30)
toyBench\CVaverageKernel.m (1407, 2010-04-23)
toyBench\benchSignaux1D.m (9530, 2010-09-01)
toyBench\SpResClass.m (2626, 2010-04-03)
toyBench\HybridWavelet.m (1975, 2010-04-01)
toyBench\SingleBest.m (3817, 2010-07-17)
toyBench\CreateInfoKernel.m (554, 2009-12-23)
toyBench\launcherperf.m (4071, 2010-08-05)
toyBench\optionsdefault.m (1136, 2010-09-01)
toyBench\CreateDataAppVal.m (304, 2009-12-23)
toyBench\setpath.m~ (1133, 2010-09-01)
toyBench\compareExhaustStockAR.m (4032, 2010-07-17)
toyBench\CVLucas.m (1186, 2010-04-19)
toyBench\launcher.m (5497, 2010-08-10)
toyBench\Exploitresults.m (7023, 2010-08-10)
toyBench\benchSignaux1DFigure.m (8959, 2010-04-30)
toyBench\plotSignalSim.m (862, 2010-04-30)
toyBench\ExploitresultsPerf.m (8002, 2010-08-10)
Comparespeed\Comparemklhessac.m (4743, 2010-07-15)
Comparespeed\mklsvmhess.m (13280, 2010-05-07)
Comparespeed\Comparemklhessac.m~ (4739, 2010-05-07)
Comparespeed\exploitresults.m~ (1405, 2010-05-12)
Comparespeed\exploitresults.m (1385, 2010-07-16)
Comparespeed\alternateWaveletK.m (393, 2010-01-27)
Comparespeed\GramMatrix3D.m (359, 2010-01-27)
Comparespeed\setpath.m (1134, 2010-04-28)
Comparespeed\setpath.m~ (1133, 2010-04-01)
Comparespeed\costhesssvmclass.m (650, 2008-10-21)
Comparespeed\gradhesssvmclass.m (1765, 2008-10-21)
Comparespeed\svqp.cpp (13744, 2008-10-14)
Comparespeed\svqp.mexglx (18318, 2008-10-14)
Comparespeed\svqp.mexa64 (20051, 2008-10-14)
WavePath.m (6206, 2009-07-29)
waveletSearchingAlgo.m (8407, 2010-08-10)
waveletProjKernel.m (2404, 2010-04-27)
waveletAtomKernel.m (6839, 2010-04-23)
wavelet2dProjKernel.m (2190, 2009-07-17)
... ...

This package should allow one to reproduce the toy results in the "wavelet kernel learning" paper. It needs our SVM-KM toolbox for training single kernel SVM. The file entries are for the toy results are : * benchSignaux1D.m which runs the comparison for the different approaches. Options can be changed in the optionsdefault.m * launcher.m allows to reproduce the results in the paper (it may take fews days of computation) * The main algorithm for the WKL algorithm in given in mklsvminf2.m * the wavelet kernel searching algorithm is in (kernel in equation 2) waveletsearchingalgo.m * the wavelet scale kernel searching is in (kernel in equation 4) scalesearchingalgo.m This package has been written by Florian Yger & Alain Rakotomamonjy, University of Rouen August 2010 and is under a GPL licence

近期下载者

相关文件


收藏者