AB2C: Artificial Bee Colony for Clustering
Seyed Mohammad Hossein Hasheminejad, Marziyeh. Vosoughian, Mohamad Zamini · 2020
Clustering is one of the most challenging unsupervised techniques. Methods that are used for clustering, are not able to meet all the needs of issues simultaneously. Clustering large datasets with many features are difficult which provide high computational complexity. Artificial bee colony algorithm (ABC) is an efficient, effective and robust technique in clustering with fewer control parameters in swarm intelligence, which emulates the foraging manner of honeybees. In this paper, the aim is to find an algorithm with suitable iteration to achieve the optimal solution. Therefore, the improved artificial bee colony algorithm, called AB2C (Artificial Bee Colony for Clustering), to converge faster has used the best solution obtained by bees to update their population and K-means algorithm to initialize the population as guided. The proposed AB2C is applied to some UCI data sets and comparing the obtained results with CPSOII and other clustering techniques such as CPSOI, CGA, and other ABC-based clustering reveals that the proposed AB2C yields promising results.