Source code of unsupervised GNTD and semi-supervised GNTD.
We provide two demo codes:
demoUGNTD.m: unsupervised GNTD for image representation.
demiSemiGNTD.m: Semi-supervised GNTD for image representation.
In this package, we used the tensor toolbox 2.6, which is downloaded from (http://www.sandia.gov/~tgkolda/TensorToolbox)
@article{qiu2020generalized,
title={A generalized graph regularized non-negative tucker decomposition framework for tensor data representation},
author={Qiu, Yuning and Zhou, Guoxu and Wang, Yanjiao and Zhang, Yu and Xie, Shengli},
journal={IEEE transactions on cybernetics},
volume={52},
number={1},
pages={594--607},
year={2020},
publisher={IEEE}
}
@inproceedings{qiu2019graph,
title={Graph regularized nonnegative tucker decomposition for tensor data representation},
author={Qiu, Yuning and Zhou, Guoxu and Zhang, Yu and Xie, Shengli},
booktitle={ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
pages={8613--8617},
year={2019},
organization={IEEE}
}