consistentgraphlearning's Introduction

Distance Graph Fusion (DGF) and Similarity Graph Fusion (SGF)

This repository contains the MATLAB code for DGF and SGF introduced in the following paper

Consistency Meets Inconsistency: A Unified Graph Learning Framework for Multi-view Clustering (ICDM 2019)

Preparation

Windows 64bit: Add some helper files to MATLAB path by addpath('MinMaxSelection') command in MATLAB command window.
Linux, Windows 32bit and Mac OS: Add some helper files to MATLAB path by addpath('MinMaxSelection') command in MATLAB command window. Then recompile the helper functions by running minmax_install.

Example usage

load('data\handwritten.mat');
knn=15; beta=1e-6; gamma=1e1;
[nmi, label] = DGF(X, Y, knn, beta, gamma);

Multi-view Data

More multi-view data are available on Google Drive.

Citation

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

@inproceedings{liang2019consistency,
  title={Consistency Meets Inconsistency: A Unified Graph Learning Framework for Multi-view Clustering},
  author={Youwei Liang, Dong Huang, and Chang-Dong Wang},
  booktitle={2019 IEEE International Conference on Data Mining (ICDM)},
  year={2019}
}

consistentgraphlearning's People

Contributors

Stargazers

Watchers

consistentgraphlearning's Issues

Can this code be executed directly?

After I downloaded this code, I tested it according to the readme.md file, but I always reported an error:: "The matrix dimensions must be consistent".

Missing function 'extract_from_idx'

Both DGF and SGF functions are missing a helper function called 'extract_from_idx' which is on lines 45 and 55 in SGF and 45 in DGF. The example provided in the README does not work without this function.

>> [nmi, label] = SGF(X, Y, knn, beta, gamma);
Undefined function or variable 'extract_from_idx'.

Error in SGF (line 45)
        [~, tp] = extract_from_idx(ONE, idx{i});