Discretize Relaxed Solution of Spectral Clustering via A Non-Heuristic Algorithm

This repository is our implementation of

Hongyuan Zhang and Xuelong Li, "Discretize Relaxed Solution of Spectral Clustering via A Non-Heuristic Algorithm," IEEE Transactions on Neural Networks and Learning Systems, DOI: 10.1109/TNNLS.2023.3309871, 2023.

In this paper, we aim to develop a non-heuristic algorithm to discretize the continuous solutions returned by the existing spectral clustering methods. The proposed algorithm is motivated by the first-order algorithm in classical optimization.

If you have issues, please email:

hyzhang98@gmail.com or hyzhang98@mail.nwpu.edu.cn.

Requirements

MATLAB $\geq$ 2019b

Citation

@article{first-order-spectral-mapping,
  author={Zhang, Hongyuan and Li, Xuelong},
  journal={IEEE Transactions on Neural Networks and Learning Systems}, 
  title={Discretize Relaxed Solution of Spectral Clustering via A Non-Heuristic Algorithm}, 
  year={2023},
  volume={},
  number={},
  pages={1--1},
  doi={10.1109/TNNLS.2023.3309871}
}