Comparison of cluster measures
Clustering algorithms can be described as unsu-pervised learning algorithms in machine learning process. They assign class labels to data objects based on the relationship between data items without any pre-defined class label. This gives rise to an inherent uncertainty in the clustering process. As such clustering validity measures are needed to verify the clusters from different clustering algorithms. In this paper we compare various cluster validity measures proposed in the literature. We use the 'usage frequency' of these measures as a distinguishing attribute to classify them. We also provide the results of our experimentation to validate our classification.
International Conference on Artificial Intelligence and Pattern Recognition 2009, AIPR 2009
Anyanwu, M., Shiva, S., & Lin, D. (2009). Comparison of cluster measures. International Conference on Artificial Intelligence and Pattern Recognition 2009, AIPR 2009, 348-355. Retrieved from https://digitalcommons.memphis.edu/facpubs/2632