An ACO-Based Clustering Algorithm With Chaotic Function Mapping

• Published in: International journal of cognitive informatics & natural intelligence Vol. 15; no. 4; pp. 1 - 21
• Main Authors: Zhang, Wensheng, Yang, Lei, Hu, Xin, Wang, Hui, Huang, Kang, Wang, Dongya
• Format: Journal Article
• Language: English
• Published: Hershey IGI Global 23.06.2022
• Subjects: Algorithms; Ant colony optimization; Big Data; Classification; Cluster analysis; Clustering; Convergence; Data mining; Datasets; Efficiency; Food; Genetic algorithms; Informatics; Mapping; Optimization; Pheromones; Sensors; China
• ISSN: 1557-3958; 1557-3966; 1557-3966
• DOI: 10.4018/IJCINI.20211001.oa20

To overcome shortcomings when the ant colony optimization clustering algorithm (ACOC) deal with the clustering problem, this paper introduces a novel ant colony optimization clustering algorithm with chaos. The main idea of the algorithm is to apply the chaotic mapping function in the two stages of ant colony optimization: pheromone initialization and pheromone update. The application of chaotic mapping function in the pheromone initialization phase can encourage ants to be distributed in as many different initial states as possible. Applying the chaotic mapping function in the pheromone update stage can add disturbance factors to the algorithm, prompting the ants to explore new paths more, avoiding premature convergence and premature convergence to suboptimal solutions. Extensive experiments on the traditional and proposed algorithms on four widely used benchmarks are conducted to investigate the performance of the new algorithm. These experiments results demonstrate the competitive efficiency, effectiveness, and stability of the proposed algorithm.