Weighted Mixed Distance (WMD) [paper]
Paper title: A Distance Metric for Clustering Mixed Data Using Graph-Based Feature Influence Balancing Approach
This repository presents the Weighted Mixed Distance (WMD) metric, an approach for clustering datasets containing both numerical and categorical features. WMD improves clustering quality by adjusting feature influence using a graph-based unsupervised method.
MATLAB code wmd(points, WR, WC) function returns
[WR, WC, LS_Table] = getweights(points);
D = wmd(points, WR, WC);
disp(D);
@INPROCEEDINGS{11167150,
author={Pyae, Aung and Low, Yeh-Ching and Chua, Hui Na},
booktitle={2025 IEEE 7th Symposium on Computers & Informatics (ISCI)},
title={A Distance Metric for Clustering Mixed Data Using Graph-Based Feature Influence Balancing Approach},
year={2025},
volume={},
number={},
pages={59-64},
keywords={Measurement;Couplings;Laplace equations;Data analysis;Clustering methods;Indexes;Informatics;Unsupervised learning;Information integrity;clustering;similarity measure;mixed data;unsupervised learning},
doi={10.1109/ISCI65687.2025.11167150}}