Hits:

Indexed by:期刊论文

Date of Publication:2014-07-15

Journal:Journal of Computational Information Systems

Included Journals:EI

Volume:10

Issue:14

Page Number:6007-6014

ISSN No.:15539105

Abstract:Data set with missing attributes is often encountered in practical applications. To solve the problem, missing attributes are represented as reconstructed intervals based on the nearest-neighbor rule in this paper. Furthermore, we propose a hybrid clustering algorithm for incomplete data set, ant colony guiding FCM based on missing attributes coding (AFA), which uses ant as a guider to search more accuracy imputations of missing attributes in the corresponding nearest-neighbor intervals. The experimental results for several UCI data sets show the superiority of the proposed global optimization algorithm. ? 2014 by Binary Information Press