/LDP-SC_code

This is the original implementation of the algorithm LDP-SC in the paper "Clustering based on local density peaks and graph cut".

Primary LanguagePythonMIT LicenseMIT

This is the original implementation of the algorithm LDP-SC in the paper "Clustering based on local density peaks and graph cut" ( DOI ; Direct Link ).

Requirements

Current environment:

  • python == 3.9
  • scipy == 1.7.2
  • numpy == 1.20.3
  • scikit-learn == 1.0.1
  • networkx == 2.6.3

Description

  • main.py -- A demo of how to use the LDP-SC algorithm.
  • main.ipynb -- Demos of loading some of the used datasets.
  • evaluation.py
    • compute_score() -- Get ARI, NMI, ACC metrics.

Citation

If you find this file useful in your research, please consider citing:

@article{long2022clustering,
  title = {Clustering based on local density peaks and graph cut},
  journal = {Information Sciences},
  volume = {600},
  pages = {263-286},
  year = {2022}
}