A new algorithm for data clustering based on cuckoo search optimization

Abstract

This paper presents a new algorithm for data clustering based on the cuckoo search optimization. Cuckoo search is generic and robust for many optimization problems and it has attractive features like easy implementation, stable convergence characteristic and good computational efficiency. The performance of the proposed algorithm was assessed on four different dataset from the UCI Machine Learning Repository and compared with well known and recent algorithms: K-means, particle swarm optimization, gravitational search algorithm, the big bang–big crunch algorithm and the black hole algorithm. The experimental results improve the power of the new method to achieve the best values for three data sets.