基于QPSO算法的支持向量机参数优化研究

摘要

应用具有量子行为的粒子群优化算法,对支持向量机(SVM)进行参数优化研究.根据支持向量机的分类准确率和泛化能力之间的关系,应用QPSO算法选取比较优秀的参数模型,比较参数模型的各项性能,选取最适合实际需要的参数模型.仿真表明,QPSO算法的SVM模型与PSO算法相比在分类准确率和泛化能力上均获得更好的效果,经QPSO优化后的SVM整体性能明显提高.%Quantum-behaved Particle Swarm Optimization (QPSO) is utilized to research parame-ter optimization of Support Vector Machine (SVM). According to the relationship between classification accuracy and generalization of SVM,better parameter models are chosen to compare their performances in order to obtain the parameter model which is the most suitable to the actual requirement. Simulation shows that QPSO can obtain the better parameter model in classification accuracy and generalization and the over-all performance of SVM has a great improvement after it has been optimized by QPSO,