进化贝叶斯优化的核极限学习机分类器  

Kernel extreme learning machine classifier based on evolutionary Bayesian optimization

作　　者：张梦蝶 覃华[1] 苏一丹[1] ZHANG Meng-die;QIN Hua;SU Yi-dan(School of Computer,Electronics and Information,Guangxi University,Nanning 530004,China)

机构地区：[1]广西大学计算机与电子信息学院,广西南宁530004

出　　处：《计算机工程与设计》2022年第2期399-405,共7页Computer Engineering and Design

基　　金：国家自然科学基金项目(51667004、61762009)。

摘　　要：为解决传统核极限学习机算法参数优化困难的问题,提高分类准确度,提出一种改进贝叶斯优化的核极限学习机算法。用樽海鞘群设计贝叶斯优化框架中获取函数的下置信界策略,提高算法的局部搜索能力和寻优能力;用这种改进的贝叶斯优化算法对核极限学习机的参数进行寻优,用最优参数构造核极限学习机分类器。在UCI真实数据集上进行仿真实验,实验结果表明,相比传统贝叶斯优化算法,所提算法能提升核极限学习机的分类精度,相较其它优化算法,所提算法可行有效。To solve the problems of parameter optimization of traditional kernel extreme learning algorithm and to improve the classification accuracy,a kernel extreme learning machine based on improved Bayesian optimization was proposed.The lower confidence bound strategy of the acquisition function in the Bayesian optimization framework was designed with the salp swarm,which improved the local search ability and searching ability of the algorithm.The improved Bayesian optimization algorithm was used to optimize the parameters of kernel extreme learning machine,and the kernel extreme learning machine classifier was constructed with the optimal parameters.Simulation experiments on UCI real data sets show that the proposed algorithm can improve the classification accuracy of kernel extreme learning machine compared with traditional Bayesian optimization algorithm,and compared with other optimization algorithms,the proposed algorithm is feasible and effective.

关 键 词：核极限学习机 核参数 贝叶斯优化 进化下置信界策略 分类精度 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]