An New Clustering Algorithm Based on QPSO and Simulated Annealing

作者：Wang Yong^*; Xu Wenbo; Sun Jun

来源：International Symposium on Distributed Computing and Applications to Business, Engineering and Science, 2008-07-27 to 2008-07-31.

• Cluster
• Simulated
• Annealing
• QPSO
• Hybridization
• Variation

Particle Swarm Optimization (PSO) algorithm is a random population-based optimization technique which is simple and effective. Quantum-behaved Particle Swarm Optimization (QPSO)is a new algorithm model based on PSO. Simulated annealing is a computational intelligence algorithm which performances great and widely used in the solving nonlinear optimization problem. Clustering in data mining is similar with simulated annealing in essence, and simulated annealing can be used in data mining and clustering analysis. This article introduces the simulation annealing thought into QPSO algorithm which merges the hybrid and Gaussian variation, and proposes a new clustering algorithm based on QPSO and simulated annealing. This algorithm keeps the characteristic of QPSO which is simple and easy to achieve, and improves the ability of global searching and raises the convergence rate and stability. The experiment result indicates that this algorithm is better than other commen clustering algorithms.

• 出版日期2008
• 单位江南大学