Efficient protocol for data clustering by fuzzy Cuckoo Optimization Algorithm

• Published in: Applied soft computing Vol. 41; pp. 15 - 21
• Main Authors: Amiri, Ehsan, Mahmoudi, Shadi
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.04.2016
• Subjects: Artificial intelligence; Cuckoo Optimization Algorithm (COA); Data clustering; Dataset; Fuzzy logic; Fuzzy logic; Cuckoo Optimization Algorithm (COA); Artificial intelligence; Dataset; Data clustering
• ISSN: 1568-4946; 1872-9681
• DOI: 10.1016/j.asoc.2015.12.008

•This paper presents a new optimization approach for data clustering with COA.•High quality results obtained for dataset.•This paper presents a new optimization approach for data clustering with COA and fuzzy system for clustering data.•An efficient proposal for data clustering by Cuckoo Optimization Algorithm.•In this proposal at each iteration, firstly generates r cuckoo's agents. Each cuckoo generates a random solution string and tries to calculate a fitness value for its solution. Data clustering is a technique for grouping similar and dissimilar data. Many clustering algorithms fail when dealing with multi-dimensional data. This paper introduces efficient methods for data clustering by Cuckoo Optimization Algorithm; called COAC and Fuzzy Cuckoo Optimization Algorithm, called FCOAC. The COA by inspire of cuckoo bird nature life tries to solve continuous problems. This algorithm clusters a large dataset to prior determined clusters numbers by this meta-heuristic algorithm and optimal the results by fuzzy logic. Firstly, the algorithm generates a random solutions equal to cuckoo population and with length dataset objects and with a cost function calculates the cost of each solution. Finally, fuzzy logic tries for the optimal solution. The performance of our algorithm is evaluated and compared with COAC, Black hole, CS, K-mean, PSO and GSA. The results show that our algorithm has better performance in comparison with them.