多元宇宙优化算法改进SVM参数
|
摘要
针对支持向量机(SVM)参数难以选择和确定的问题,采用一种新式元启发式优化算法——多元宇宙优化算法(MVO).并在传统多元宇宙优化算法(MVO)的基础上针对TDR值下降速度慢而导致旅行距离增加的问题,提出改进多元宇宙优化算法(IMVO),将改进多元宇宙优化算法用于支持向量机的参数优化和选择问题上.使用UCI标准数据库中的数据进行数值仿真实验.研究结果表明:采用改进多元宇宙优化算法优化的支持向量机有较强的寻优性能,稳定性较好.
In order to solve the support vector machine(SVM) parameters is difficult to select and determine,this paper proposed a new meta-heuristie algorithm—Multi-verse optimizer algorithm(MVO). For TDR reduced speed slow and cause travel distance increase, an improving multi-verse optimizer algorithm(IMVO) was proposed, which is used to optimize and choose the parameters of Support Vector Machine(SVM), and the numerical simulation experiment was conducted with the datasets from University of California Irvine(UCI). The results show that the IMVO algorithm for SVM parameter optimization has stronger optimization performance and better stability.
In order to solve the support vector machine(SVM) parameters is difficult to select and determine,this paper proposed a new meta-heuristie algorithm—Multi-verse optimizer algorithm(MVO). For TDR reduced speed slow and cause travel distance increase, an improving multi-verse optimizer algorithm(IMVO) was proposed, which is used to optimize and choose the parameters of Support Vector Machine(SVM), and the numerical simulation experiment was conducted with the datasets from University of California Irvine(UCI). The results show that the IMVO algorithm for SVM parameter optimization has stronger optimization performance and better stability.
引文
[1]MIRJALILI S,MIRJALILI S M,HATAMLOU A.Multi-Verse optimizer:a nature inspired algorithm for global optimization[J].Neural Computing and Applications,2016,2(27):495-513.
[2]邓乃扬,田英杰.支持向量机[M].北京:科学出版社,2009:43-105.
[3]VAPNIK V N,VAPNIK V.Statistical Learning Theory[M].New York:Wiley,1998.
[4]邵良杉,马寒,温廷新.因子分析与支持向量机相结合的煤炭自燃预测[J].辽宁工程技术大学学报(自然科学版),2014,33(4):433-436.doi:10.3969/j.issn.1008-0562.2014.04.001SHAO Liangshan,MA han,WEN Tingxin.Coal spontaneous combustion prediction model of support vector machine combined with factor analysis[J].Journal of Liaoning Technical University(Natural Science),2014,33(4):433-436.doi:10.3969/j.issn.1008-0562.2014.04.001
[5]崔东文.支持向量机在湖库营养状态识别中的应用[J].水资源保护,2013,29(4):26-30.CUI Dongwen.The application of support vector machine to lake and reservoir trophicstate recognition[J].Water Resources Procetion,2013,29(4):26-30.
[6]董春曦,饶鲜,杨绍全,等.支持向量机参数选择方法研究[J].系统工程与电子技术,2004,26(8):1 117-1 120.DONG Chunxi,YAO Xian,YANG Shaoquan,et al.Method for selecting the parameters of support vector machines[J].Systems Engineering and Electronics,2004,26(8):1 117-1 120.
[7]刘昌平,范明钰,王光卫,等.基于梯度算法的支持向量机参数优化方法[J].控制与决策,2008,23(11):1 291-1 295.LIU Changping,FAN Mingyu,WANG Guangwei,et al.Optimizing parameters of support vector machine based on gradient algorithm[J].Control and Decision,2008,23(11):1 291-1 295.
[8]邵信光,杨慧中,陈刚.基于粒子群优化算法的支持向量机参数选择及其应用[J].控制理论与应用,2007,23(5):740-743.SHAO Xinguang,YANG Huizhong,CHEN Gang.Parameters seleetion and application of support vector maehines based on Partiele swarm optimization algorithm[J].Control Theory and Application,2007,23(5):740-743.
[9]柴岩,周艳钊.遗传算法的爬山法改进[J].辽宁工程技术大学学报(自然科学版),2014,33(7):996-999.doi:10.3969/j.issn.1008-0562.2014.07.CHAI Yan,ZHOU Yanzhao.Improved genetic algorithm based on climbing[J].Journal of Liaoning Technical University(Natural Science),2014,33(7):996-999.doi:10.3969/j.issn.1008-0562.2014.07.028
[10]杨柳松,何光宇.基于改进粒子群优化的SVM故障诊断方法[J].计算机工程,2013,39(3):187-190.YANG Liusong,HE Guangyu.Support vector machine fault diagnosis method based on improved particle swarm optimization[J].Computer Engineering,2013,39(3):187-190.
[11]GAO L,ZHAO S,GAO J.Application of artificial fishswarm algorithm in SVM parameter optimization selection[J].Computer Engineering and Applications,2013(23):2.
[2]邓乃扬,田英杰.支持向量机[M].北京:科学出版社,2009:43-105.
[3]VAPNIK V N,VAPNIK V.Statistical Learning Theory[M].New York:Wiley,1998.
[4]邵良杉,马寒,温廷新.因子分析与支持向量机相结合的煤炭自燃预测[J].辽宁工程技术大学学报(自然科学版),2014,33(4):433-436.doi:10.3969/j.issn.1008-0562.2014.04.001SHAO Liangshan,MA han,WEN Tingxin.Coal spontaneous combustion prediction model of support vector machine combined with factor analysis[J].Journal of Liaoning Technical University(Natural Science),2014,33(4):433-436.doi:10.3969/j.issn.1008-0562.2014.04.001
[5]崔东文.支持向量机在湖库营养状态识别中的应用[J].水资源保护,2013,29(4):26-30.CUI Dongwen.The application of support vector machine to lake and reservoir trophicstate recognition[J].Water Resources Procetion,2013,29(4):26-30.
[6]董春曦,饶鲜,杨绍全,等.支持向量机参数选择方法研究[J].系统工程与电子技术,2004,26(8):1 117-1 120.DONG Chunxi,YAO Xian,YANG Shaoquan,et al.Method for selecting the parameters of support vector machines[J].Systems Engineering and Electronics,2004,26(8):1 117-1 120.
[7]刘昌平,范明钰,王光卫,等.基于梯度算法的支持向量机参数优化方法[J].控制与决策,2008,23(11):1 291-1 295.LIU Changping,FAN Mingyu,WANG Guangwei,et al.Optimizing parameters of support vector machine based on gradient algorithm[J].Control and Decision,2008,23(11):1 291-1 295.
[8]邵信光,杨慧中,陈刚.基于粒子群优化算法的支持向量机参数选择及其应用[J].控制理论与应用,2007,23(5):740-743.SHAO Xinguang,YANG Huizhong,CHEN Gang.Parameters seleetion and application of support vector maehines based on Partiele swarm optimization algorithm[J].Control Theory and Application,2007,23(5):740-743.
[9]柴岩,周艳钊.遗传算法的爬山法改进[J].辽宁工程技术大学学报(自然科学版),2014,33(7):996-999.doi:10.3969/j.issn.1008-0562.2014.07.CHAI Yan,ZHOU Yanzhao.Improved genetic algorithm based on climbing[J].Journal of Liaoning Technical University(Natural Science),2014,33(7):996-999.doi:10.3969/j.issn.1008-0562.2014.07.028
[10]杨柳松,何光宇.基于改进粒子群优化的SVM故障诊断方法[J].计算机工程,2013,39(3):187-190.YANG Liusong,HE Guangyu.Support vector machine fault diagnosis method based on improved particle swarm optimization[J].Computer Engineering,2013,39(3):187-190.
[11]GAO L,ZHAO S,GAO J.Application of artificial fishswarm algorithm in SVM parameter optimization selection[J].Computer Engineering and Applications,2013(23):2.