Computer Engineering and Applications ›› 2011, Vol. 47 ›› Issue (10): 38-40.

• 研究、探讨 • Previous Articles     Next Articles

Improved PSO and its application to SVM parameter optimization

CHEN Zhiming   

  1. Department of Electronic Science,Huizhou University,Huizhou,Guangdong 516007,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-04-01 Published:2011-04-01



  1. 惠州学院 电子科学系,广东 惠州 516007

Abstract: Support Vector Machine(SVM) is a machine learning algorithm with good performance.The selection of its parameters has a great influence on its modeling accuracy and generalization performance,and it is an important area in machine learning research.Based on a brief introduction to basic Particle Swarm Optimization(PSO),a new quantum PSO algorithm is presented,as well as its implementation.Performance comparison with classic PSO algorithm is made through 4 benchmark test functions.Based on the proposed quantum PSO,the optimum parameter selection of Least Squares SVM(LS-SVM) is studied.Simulation results show that the presented quantum PSO algorithm can achieve good performance.

Key words: quantum Particle Swarm Optimization(PSO), Least Squares Support Vector Machine(LS-SVM), benchmark test, parameter optimization

摘要: 支持向量机是一种性能优越的机器学习算法,而其参数的选择对建模精度和泛化性能等有着重要的影响,也是目前机器学习研究的一个重要方向。在简要介绍基本粒子群优化(PSO)算法的基础上,提出了一种量子粒子群优化算法,给出了其实现方式,并通过4个基准测试函数进行性能对比评价。基于这种量子粒子群优化算法,对最小二乘支持向量机(LS-SVM)的参数优化进行了研究。仿真结果表明,量子粒子群优化算法能给出很好的优化结果。

关键词: 量子粒子群, 最小二乘支持向量机, 基准测试, 参数优化