The application of Particle Swarm Optimization in Training Support Vector Machines
Large number of example vectors brings difficulties for quadratic programming problem with support vector machines, traditional methods may be impossible. The intelligent search technologies, such as genetic algorithms and particle swarm optimization algorithm, can give a similar solve of problems in less time. Particle Swarm Optimization is better than genetic algorithms in convergence and stability of the overall. According to the characters of swarm intelligence and constrained optimization, we propose a method to solve a linearly constrained quadratic optimization problem in training support vector machines with PSO (for short). Testify PSO has determinate applied value in the field of support vector machines, and it is a new way for quadratic programming problem with a large number of example vectors.
support vector machines Quantum-behaved Particle Swarm Optimization quadratic programming problem
Zhiguo Chen Yan Shan
School of Information Technology, Southern Yangtze University, Wuxi, Jiangsu, 214122,China
国际会议
杭州
英文
715-718
2006-10-12（万方平台首次上网日期，不代表论文的发表时间）