摘要
针对粒子群算法优化SVM模型参数在进化后期容易陷入局部最优的问题,研究了细菌觅食趋利避害机制,提出了一种基于细菌觅食特性改进粒子群算法的方法,并将改进方法应用于优化SVM预测模型参数的研究;实验结果表明,该方法能够弥补粒子群算法在进化后期容易陷入局部最优的缺陷,具备更好的寻优性能。
In order to solve the problem that the basic particle swarm algorithm to optimize the SVM model is easy to tall into local opti mum in the late evolution, on the basis of studying mechanism of bacterial foraging theory, this paper proposes a method to improve the basic particle swarm by using bacterial foraging features, and applied it to optimize the parameters of SVM prediction model. The results show that it can remedy the defect of the basic particle swarm algorithm and has better optimization performance.
出处
《计算机测量与控制》
北大核心
2014年第6期1902-1904,共3页
Computer Measurement &Control
关键词
细菌觅食特征
粒子群算法
支持向量机
故障预测
bacterial foraging features
particle swarm algorithm
support vector machine
fault prediction
