|
Home Technical Report Software Datasets Demo ManifoldLearn Matlab Routines for Manifold RegularizationContact: Vikas Sindhwani (vikass@cs.uchicago.edu)ManifoldLearn
implements Laplacian Regularized Least Squares
and Laplacian Support Vector Machines
algorithms. Clustering,
semi-supervised and fully supervised learning tasks can be handled within
the same conceptual and algorithmic framework.
This software was tested with matlab 6.5 R13 running under Linux.
Also tested under windows running cygwin (Put cygwin1.dll in your windows
path or invoke matlab from cygwin;
mex files were compiled under cygwin using gnumex ). Usage:
Here are some examples to get started. Type "help
filename" for more information.
Generate an options structure for training options=ml_options('Kernel','rbf','KernelParam',0.5,'NN',6); Train a Classifier classifier=ml_train(X,Y,options,'lapsvm'); Test a classifier [real_output, binary_labels, error_rate]=ml_test(classifier, Xtest, Ytest); Computing Kernels K=calckernel('rbf', 1, X); calckernel computes linear, polynomial or RBF kernels. Computing Graphs Laplacian L=laplacian(X,'nn',options) Graphs can be constructed with euclidean or cosine distance measures; with binary weights or heat kernel weights; normalized or unnormalized. Training Algorithms Implemented in svm.m, rlsc.m, lapsvm.m, laprlsc.m -- these routines take gram matrix and graph laplacian matrix, and output expansion coefficients (alphas). A mex interface to libsvm is used for SVM QP. Evaluating Classifiers Can use ROC curves, precision-recall breakeven points, etc. Uses some of Jason Rennie's Matlab Code.
|