- Avrupa Bilim ve Teknoloji Dergisi
- Vol: 5 Issue: 9
- A New Clustering Algorithm of Hybrid Data According to Weights of Attributes
A New Clustering Algorithm of Hybrid Data According to Weights of Attributes
Authors : Osman Çörekci, Ayla Sayli
Pages : 28-37
View : 19 | Download : 6
Publication Date : 2016-12-26
Article Type : Research
Abstract :Separating large data into similar clusters is one of the basic problems of data mining. Storing large data in an organized way has currently increased the importance of the methods developed for clustering. Even if the hierarchical clustering methods give effective results, they are still inadequate due to their computational complexity. Non-hierarchical clustering methods cannot be used for all data types because of the cost function which cannot run by categorical data. Recently, some non-hierarchical clustering methods have been improved for categorical and hybrid data. In addition, the weights of attributes in clustering might be different due to the nature of the data or the expected results. In this paper, we introduce an algorithm which has been improved for the clustering of large hybrid data in an effective way that also includes the weights of attributes. This algorithm, mainly based on the K-Prototypes algorithm, will be called "W-K-Prototypes”. The computational results show that the algorithm can be used efficiently for clustering.Keywords : Data mining, clustering, clustering analysis, clustering methods, clustering hybrid data, K-Prototypes algorithm, weighted attributes