Auto-Evolving Clusters based on Rejection and Migration
Jyotı Lakhani, Ajay Khunteta, Anupama Chowdhary, Dharmesh Harwani · 2016
In the present communication an auto evolving evolutionary clustering Algorithm, Auto-Evolving Clusters based on Rejection And Migration (AEC-RAM) is being introduced. We have given an attempt to develop an algorithm to overcome the disparity of K-means in order to understand the number of clusters a-priori. The algorithm starts with considering all items in a data set as a single cluster. Migration operator is used to auto evolve the cluster and to find out the remaining clusters in a given dataset. Partitioning of clusters continues until the total individual items get adapted to their clusters and subsequently the population will get stable. The algorithm is tested with UCI datasets and produces relatively better results.