This paper discusses a density based clustering approach for a guided kernel based clustering algorithm, named MK-means (Modified K-means). Our idea is to improve the guided K-Means clustering algorithm and discuss the benefits of using MK-Means algorithm for clustering algorithm in astrophysics data bases. The improvements made allow handling clustering without apriori knowledge and also include the flexibility of merging classes when similarities are detected.
|Title of host publication||IEEE International Joint Conference on Neural Networks (IJCNN)|
|Publication status||Published - 1 Jan 2010|
|Event||IEEE World Congress on Computational Intelligence (WCCI 2010) - |
Duration: 1 Jan 2010 → …
|Conference||IEEE World Congress on Computational Intelligence (WCCI 2010)|
|Period||1/01/10 → …|