Manifold-based Constraint Laplacian Score for multi-label feature selection
This code from our paper "Manifold-based Constraint Laplacian Score for multi-label feature selection".
function idx = MCLS(fea, knear)
Inputs:
fea: struct data set
data : nSample * nFeature
target : nLabel * nSample
knear: number of neighbors
Output:
idx: feature sorting by score
Usage:
fea = load('sample.mat');
knear = 5;
idx = MCLS(fea,knear);