资 源 简 介
CARS-PLSLDA
The updated version is available only at: http://www.libpls.net
Currently, CARS coupled with partial least squares linear discriminant analysis(PLSLDA), is an efficient strategy for identiying an optimal subset of variales for classification. An exponentially decreasing function(EDF) is introduced into CARS for highly efficient screening of variables, which is especially suitable for (very)high dimensional data, e.g. gene expression profile. CARS is short for Competitive Adaptive Reweighted Sampling. Please refer to Li, Liang, Xu and Cao, Anal. Chim. Acta, 2009, 648 (1): 77-84 for details. Note that considering the reproducibility of CARS, we have removed Monte Carlo operation of the original CARS presented in the literature. Therefore, simplified version of CARS is implemented here for the sake of exactly reproducing the variable selection result with the same parametric settings.
http://cars
文 件 列 表
CARS_PLSLDA
carsplslda.m
classplot2.m
classplot2random.m
classpoint.m
compareplot.m
confounderajust.m
COSS.jpg
csvd.m
databin.m
Distribution.jpg
DM2.mat
example_spa.m
fishersel.m
histfitnew.m
ks.m
ldapinv.m
LOGO_CARS.JPG
mcuveplslda.m
msst.m
mtc.m
optisim.m
plot_train_test.m
plotcars.m
plotlda.m
plotpra.m
plotspa.m
PLS-LDA_Manual.doc
pls_nipals.m
plslda.m
plsldacv.m
plsldadcv.m
plsldamccv.m
plsldaproj.m
plsldardcv.m
plsldaval.m
plssim.m
pra.m
pretreat.m
rmvconstvar.m
roccurve.m
scarsplslda.m
sesp.m
spa.m
SPA_manual.doc
splineValue.m
test_package_functions.m
tp.m
traintestselect.m
tstatis.m
tvalue.m
VariablesWeight.m
vipp.m
weightsampling.m
wsample.m
