A new approach for weighted clustering using decision tree

In the field of cluster analysis, most clustering algorithms consider the contribution of each attribute for classification uniformly. In fact, different attributes (or different features) should be of different contribution for clustering result. In order to consider the different roles of each att...

Full description

Saved in:
Bibliographic Details
Main Authors: Dogan, Y., Birant, D., Kut, A.
Format: Conference Proceeding
Language:English
Subjects:
Algorithm design and analysis
Clustering algorithms
decision tree
Decision trees
Diabetes
Euclidean distance
Glass
Iris
k-means
machine learning
rule set
Weighted clustering
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In the field of cluster analysis, most clustering algorithms consider the contribution of each attribute for classification uniformly. In fact, different attributes (or different features) should be of different contribution for clustering result. In order to consider the different roles of each attribute, this paper proposes a new approach for clustering algorithms based on weights, in which decision tree technique is used to assign the weights to each attribute. The comparison results show that novel approach improves the robustness of the traditional clustering algorithms. The experimental results with various test data sets illustrate the effectiveness of the proposed approach.
DOI:10.1109/INISTA.2011.5946126