An empirical comparison of four initialization methods for the K-Means algorithm
Pattern Recognition Letters
Rough Sets: Theoretical Aspects of Reasoning about Data
Rough Sets: Theoretical Aspects of Reasoning about Data
On Clustering Validation Techniques
Journal of Intelligent Information Systems
Interval Set Clustering of Web Users with Rough K-Means
Journal of Intelligent Information Systems
Cluster center initialization algorithm for K-means clustering
Pattern Recognition Letters
Some refinements of rough k-means clustering
Pattern Recognition
A method for initialising the K-means clustering algorithm using kd-trees
Pattern Recognition Letters
k-means++: the advantages of careful seeding
SODA '07 Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms
Rough Cluster Quality Index Based on Decision Theory
IEEE Transactions on Knowledge and Data Engineering
A reasonable rough approximation for clustering web users
WImBI'06 Proceedings of the 1st WICI international conference on Web intelligence meets brain informatics
K-means clustering seeds initialization based on centrality, sparsity, and isotropy
IDEAL'09 Proceedings of the 10th international conference on Intelligent data engineering and automated learning
| Hi-index | 0.00 |
In this paper, we present the discernibility approach for an initial seed computation of Rough K-Means (RKM). We propose the use of the discernibility initial seed computation (ISC) for RKM. Our proposed algorithm aims to improve the performance and to avoid the problem of an empty cluster which affects the numerical stability since there are data constellations where |Ck| = 0 in RKM algorithm. For verification, our proposed algorithm was tested using 8 UCI datasets and validated using the David Bouldin Index. The experimental results showed that the proposed algorithm of the discernibility initial seed computation of RKM was appropriate to avoid the empty cluster and capable of improving the performance of RKM.