Implementation of Hard C-Means Clustering

Date: 30 August, 2012
Location: New Delhi
Team Size: 1
Details:
This project was done after Fuzzy C-Means project to help the first-year students of the junior M.Sc. batch of 2012-2014 from Department of Computer Science, University of Delhi understand the concept of Hard C-Means Clustering and help them understand MATLAB programming. It uses MATLAB to perform a Hard C-Means Clustering on a dataset of less than ten data points and two attributes, and creates two clusters, though this can be changed depending on the dataset used. A fixed number of clusters are created and the data points are allotted to clusters with a membership coefficient. The clusters centers are updated till there is virtually no change between two consecutive clusters.

MATLAB code to perform the Hard C-Means (HCM) Clustering and the datasets used in the program can be found here.