Supervised and constrained Non-negative Matrix Factorization with sparseness
This method is used for the binary and multi-classification problem on images, which can make the data belong to the one class collapse to a positon in the subspace and maintain sparsity.
This code is a reproduce code from the original paper Supervised and constrained nonnegative matrix factorization with sparseness for image representation
Cite page:
@article{cai2018supervised,
title={Supervised and constrained nonnegative matrix factorization with sparseness for image representation},
author={Cai, Xibiao and Sun, Fuming},
journal={Wireless Personal Communications},
volume={102},
number={4},
pages={3055--3066},
year={2018},
publisher={Springer}
}