Spider Extra Algorithms
To install these algorithms please unzip into the same directory where
spider is installed. The zip files contain the right target subdirectory.
You can also download a snapshot of all extra algorithms here.
Core library objects
wilcoxon
Wilcoxon object - significance test of results
joint_kernel
Joint kernel object - for calculating inner
products in joint feature spaces
joint_data
Joint_data object
corrt_test
corrected resampled t-test object - significance
test of results
Unsupervised objects
spectral
Spectral clustering
ppca
Probabilistic Principal Components Analysis
nmf
Non-negative Matrix Factorization object
mrank
Manifold Ranking
Density Estimation objects
indep
Feature selection by independent density
estimation
Feature Selection objects
r2w2_sel
Feature scaling/selection via SVMs and r^2/w^2
bound.
nfe
Non-Linear Feature Elimination
mutinf
Feature selection by mutual information
mars
Greedy selection algorithm by J.Friedmann
l0
L0 zero-norm minimization (Weston,Elisseeff dual
method) object
fsv
Primal zero-norm based feature selection by
O.Mangasarian
feat_sel
FEAT_SEL feat_sel object
fisher
Select features based on Fisher/Correlation
score
clustub
Multi-class feature selection using spectral
clustering
Multi-Class and Multi-label objects.