CVDD

Version 1.0 (827 ko) par Lianyu Hu
An Internal Validity Index Based on Density-Involved Distance
0.0
(0)
24 téléchargements
Mise à jour 22 oct. 2021
Afficher la licence sur GitHub

An Internal Validity Index Based on Density-Involved Distance

https://ieeexplore.ieee.org/document/8672850

PWC PWC PWC PWC PWC

Requirement

The source code is written by Matlab r2016a. Versions lower than Matlab r2012a have not been tested. View CVDD on File Exchange

Simplest Demo

  • RUN Ncut_test.m to determine the optimal partition from varied partitions (produced by Jianbo Shi's Normalized cuts).

    CVDD.m includes Algorithm 1: CVDD in our paper.

    Ncut_test.m as an example includes Algorithm 2: CVDD-OP in our paper.

    'OP_CA' (in Ncut_test.m) shows the comparison. [ Purity, CVDD, CVNN, WB, Silhouette, CH, DB, Dunn, S_Dbw, I ]

Parameters in CVDD

No need to tune.

Datasets used

File Datasets_all30 includes 10 non-spherical clusters, 10 spherical clusters and 10 classification datasets (real datasets) used in the experiments of our paper.

Issues, Questions, etc

Please report issues here on the github page or contact "hly4ml@gmail.com"

Citation pour cette source

Lianyu Hu (2024). CVDD (https://github.com/hulianyu/CVDD/releases/tag/v1.0), GitHub. Récupéré le .

Hu, Lianyu, and Caiming Zhong. An Internal Validity Index Based on Density-Involved Distance. Institute of Electrical and Electronics Engineers (IEEE), 2019, pp. 40038–51, doi:10.1109/access.2019.2906949.

Afficher d’autres styles
Compatibilité avec les versions de MATLAB
Créé avec R2016a
Compatible avec les versions R2012a et ultérieures
Plateformes compatibles
Windows macOS Linux

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Ncut

RS

Version Publié le Notes de version
1.0

Pour consulter ou signaler des problèmes liés à ce module complémentaire GitHub, accédez au dépôt GitHub.
Pour consulter ou signaler des problèmes liés à ce module complémentaire GitHub, accédez au dépôt GitHub.