Data Mining and Knowledge Engineering

Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Selecting Optimal Weighted Medoids for Clustering


Affiliations
1 Department of Computer Science & Engineering, R.V.R. & J.C. College of Engineering, Chowdavaram, Guntur, A.P., India
2 Department of Computer Science & System Engineering, Andhra University College of Engineering, Andhra University, Visakhapatnam, A.P., India
     

   Subscribe/Renew Journal


Clustering is the process of grouping data into clusters. Partitioning is an important clustering method. K-medoids is a classical partitioning method. K-medoids generates k clusters for a dataset of n objects using distance measure. But this algorithm may not be suitable for many real life applications. In addition to distance measure, the medoid selection may depend on many other factors in real life. A new weighted k-medoids algorithm is proposed in this paper to find optimal medoids using a new measure, weights of the medoids.

Keywords

K-Medoids, Weighted K-Medoids, Partitioning Clustering, SSE.
User
Subscription Login to verify subscription
Notifications
Font Size

Abstract Views: 264

PDF Views: 2




  • Selecting Optimal Weighted Medoids for Clustering

Abstract Views: 264  |  PDF Views: 2

Authors

B. Vara Prasada Rao
Department of Computer Science & Engineering, R.V.R. & J.C. College of Engineering, Chowdavaram, Guntur, A.P., India
M. Sreelatha
Department of Computer Science & Engineering, R.V.R. & J.C. College of Engineering, Chowdavaram, Guntur, A.P., India
M. Shashi
Department of Computer Science & System Engineering, Andhra University College of Engineering, Andhra University, Visakhapatnam, A.P., India

Abstract


Clustering is the process of grouping data into clusters. Partitioning is an important clustering method. K-medoids is a classical partitioning method. K-medoids generates k clusters for a dataset of n objects using distance measure. But this algorithm may not be suitable for many real life applications. In addition to distance measure, the medoid selection may depend on many other factors in real life. A new weighted k-medoids algorithm is proposed in this paper to find optimal medoids using a new measure, weights of the medoids.

Keywords


K-Medoids, Weighted K-Medoids, Partitioning Clustering, SSE.