Improved the Performance of Cluster Oriented Classifier using Ant Colony Optimization (AECC)
S C Mishra, Anurag Kumar Jain · 2013
The generation and selection of optimal cluster for cluster based ensemble classifier is important parameter. The selection of optimal cluster impact the performance of ensemble classifier, in conventional cluster based ensemble classifier used cluster diversity such as hard clustering with agglomerative cluster and some other cluster technique. The diversity of clustering technique used fixed number of cluster selection due to fixed number of generated cluster. For the improvement of optimal cluster selection used ANT colony optimizations technique for generation of multiple cluster index and cluster confidence value. The multiple cluster confidence value gives the better selection of optimal cluster selection process. In this paper proposed ant selection based cluster ensemble classifier for data classification. Empirical evaluation shows better result in compression of COEC.