A research and application of continuous attributes discretization based on cloud model and information entropy

Authors:
Qiaoyun Huang;Guolong Chen;Yanhua Liu;Wenzhong Guo;Xiaotong Fang
Affiliations:
College of Mathematics and Computer Science, Fuzhou University, Fuzhou, China;College of Mathematics and Computer Science, Fuzhou University, Fuzhou, China and Key Laboratory of Discrete Mathematics with Application of Ministry of Education, Fuzhou, China;College of Mathematics and Computer Science, Fuzhou University, Fuzhou, China;College of Mathematics and Computer Science, Fuzhou University, Fuzhou, China;College of Mathematics and Computer Science, Fuzhou University, Fuzhou, China
Venue:
FSKD'09 Proceedings of the 6th international conference on Fuzzy systems and knowledge discovery - Volume 3
Year:
2009

Citing 3
Cited 0

On the Handling of Continuous-Valued Attributes in Decision Tree Generation

Machine Learning
Mining quantitative association rules in large relational tables

SIGMOD '96 Proceedings of the 1996 ACM SIGMOD international conference on Management of data
DrC4.5: Improving C4.5 by means of prior knowledge

Proceedings of the 2005 ACM symposium on Applied computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Discretization of continuous attributes is one of the most important issues in network data preprocessing. In this paper, a discretization algorithm of continuous attributes based on cloud model and information entropy is proposed. Making use of cloud transform, the proposed algorithm partitions the domain of every continuous attribute into many concepts represented by cloud models. The uncertain boundary of cloud model is more appropriate for actual data distribution. Define information entropy for every candidate cloud that treated as a measuring of importance. On the basis of that, select appropriate neighboring concepts and merge them. It could increase the information granularity of information system. Then utilize the fuzziness of cloud model for realizing the adaptive adjustment of boundary data, so as to improve consistency level in decision table. By employing the new algorithm, the experiments on Iris data set reached the expectation. The results show that the algorithm is feasible and efficient.