Investigation of the performance of some clustering algorithms.
Tin Zar Kyaw.
Date of Issue2005
School of Electrical and Electronic Engineering
Abstract In this dissertation, we investigate the performance of K-means, SOM and EM clustering algorithms based on UCI datasets. Tests are performed using three datasets. Two measures which are mostly used for comparing the performance of clustering algorithms. The first concerns the quality of clustering algorithms. The second measure deals with the accuracy of clustering results. The results of the experiments suggest that Expectation Maximization (EM) is more robust to outliners than K-means and SOM.
DRNTU::Engineering::Electrical and electronic engineering::Computer hardware, software and systems
Nanyang Technological University