/
performBG.m
19 lines (17 loc) · 1.11 KB
/
performBG.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% This is a demo for the MDEC-HC, MDEC-SC, and MDEC-BG algorithms, %
% which are proposed in the following paper: %
% %
% D. Huang, C.-D. Wang, J.-H. Lai, and C.-K. Kwoh. %
% "Toward Multi-Diversified Ensemble Clustering of High-Dimensional %
% Data: From Subspaces to Metrics and Beyond". %
% IEEE Transactions on Cybernetics, 2022, 52(11), pp.12231-12244. %
% DOI: https://doi.org/10.1109/TCYB.2021.3049633 %
% %
% The code has been tested in Matlab R2016a and Matlab R2016b. %
% GigHub: https://github.com/huangdonghere/MDEC %
% Written by Huang Dong. (huangdonghere@gmail.com) %
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function label = performBG(baseClsSegs, ECI, clsNum)
weightedB = bsxfun(@times, baseClsSegs, ECI); clear baseClsSegs
label = Tcut(weightedB',clsNum);