ISSN : 2319-7323
INTERNATIONAL JOURNAL OF COMPUTER SCIENCE ENGINEERING |
|
|
ABSTRACT
Title |
: |
Double Centers based Efficient Initial Clusters for K-Means |
Authors |
: |
Aarti Chaudhary, Dharmveer Singh Rajpoot |
Keywords |
: |
Cluster; initial center; dataset; validity measure; k-means. |
Issue Date |
: |
September 2015 |
Abstract |
: |
Now a days, K-Means is one of the widely used algorithm for partitioning the data into clusters. The main advantage is easy to use and understand. By seeing the drawback side, the random selection of clustering initial centers will not always the optimal clustering structure. To acquire optimal clustering the initial centers selection is also efficient one. In the proposed method by using twice operations of K-means would give us good start of clustering. From the experiments we can show that intra-cluster i.e. similarity within the clusters is maximized and inter-cluster i.e. dissimilarity between the clusters is minimized which proves our propose algorithm choosing the initial centers effectively. |
Page(s) |
: |
237-243 |
ISSN |
: |
2319-7323 |
Source |
: |
Vol. 4, No.5 |
|
|
|