一种自适应子融合集成多分类器方法
|
摘要
融合集成方法已经广泛应用在模式识别领域,然而一些基分类器实时性能稳定性较差,导致多分类器融合性能差,针对上述问题提出了一种新的基于多分类器的子融合集成分类器系统;该方法考虑在度量层融合层次之上通过对各类基多分类器进行动态选择,票数最多的类别作为融合系统中对特征向量识别的类别,构成一种新的自适应子融合集成分类器方法;实验表明,该方法比传统的分类器以及分类融合方法识别准确率明显更高,具有更好的鲁棒性。
Fusion integration method has been widely used in the field of pattern recognition.However,some base classifiers have poor real-time performance stability,which causes poor performance of multiple classifiers.A new multi-classifier-based subfusion integration classification is proposed for the above problems.This method considers the dynamic selection of various classifiers at the level of measurement layer fusion,the category with the highest number of votes is the category identified by the feature vector in the fusion system to constitute a new adaptive sub-fusion integration classifier method.Experiments show that this method is significantly more accurate than conventional classifiers and classification fusion methods and has better robustness.
Fusion integration method has been widely used in the field of pattern recognition.However,some base classifiers have poor real-time performance stability,which causes poor performance of multiple classifiers.A new multi-classifier-based subfusion integration classification is proposed for the above problems.This method considers the dynamic selection of various classifiers at the level of measurement layer fusion,the category with the highest number of votes is the category identified by the feature vector in the fusion system to constitute a new adaptive sub-fusion integration classifier method.Experiments show that this method is significantly more accurate than conventional classifiers and classification fusion methods and has better robustness.
引文
[1]李艳秋,任福继,胡敏.动态模糊密度的多分类器融合算法[J].电子学报,2018,46(5):1246-1252.
[2]周星,刁兴春,曹建军,等.基于重采样和集成选择的适用于实体识别的多分类器系统[J].数字采集与处理,2017(5):931-938.
[3]贾澎涛,李阳.基于选择性集成分类器的面部表情识别研究[J].计算机应用研究,2017,34(12):3825-3827.
[4]楚浩宇,高萌,刘永生.基于并行组合分类器的脱机手写体数字识别[J].计算机技术与发展,2018(3):105-108.
[5]郝红卫,王志彬,殷绪成,等.分类器的动态选择与循环集成方法[J].自动化学报,2011,37(11):1290-1295.
[6]Zia M S,Hussain M,Jaffar M A.A novel spontaneous facial expression recognition using dynamically weighted majority voting based ensemble classifier[J].Multimedia Tools and Applications,2018,77(19):25537-25567.
[7]唐彪,金炜,符冉迪,等.多稀疏表示分类器决策融合的人脸识别[J].电信科学,2018(4):31-40.
[8]张扬,杨建华,侯宏.一种基于证据理论的数据分聚类融合算法[J].科学技术与工程,2018(1):54-58.
[9]王江,孙美凤,张炜.基于行为知识空间的多分类器网络流量分类方法[J].扬州大学学报:自然科学版,2016(4):54-57.
[10]袁立,穆志纯.基于子分类器融合的部分遮挡人耳识别[J].仪器仪表学报,2011,32(1):186-193.
[11]朱波,陈科,徐君,等.平均分布集成策略:一种新的分类器融合方法[J].小型微型计算机系统,2016,(7):1546-1550.
[12]Borja S P,et al.Ensemble feature selection for rankings of features[J].Springer International Publishing,2015,9095:29-42.
[13]陶晓玲,亢蕊楠,刘丽燕.基于选择性集成的并行多分类器融合方法[J].计算机工程与科学,2018(5):787-792.
[14]刘明,袁保宗,苗振江,等.从局部分类精度到分类置信度的变换[J].计算机研究与发展,2008,45(9):1612-1619.
[2]周星,刁兴春,曹建军,等.基于重采样和集成选择的适用于实体识别的多分类器系统[J].数字采集与处理,2017(5):931-938.
[3]贾澎涛,李阳.基于选择性集成分类器的面部表情识别研究[J].计算机应用研究,2017,34(12):3825-3827.
[4]楚浩宇,高萌,刘永生.基于并行组合分类器的脱机手写体数字识别[J].计算机技术与发展,2018(3):105-108.
[5]郝红卫,王志彬,殷绪成,等.分类器的动态选择与循环集成方法[J].自动化学报,2011,37(11):1290-1295.
[6]Zia M S,Hussain M,Jaffar M A.A novel spontaneous facial expression recognition using dynamically weighted majority voting based ensemble classifier[J].Multimedia Tools and Applications,2018,77(19):25537-25567.
[7]唐彪,金炜,符冉迪,等.多稀疏表示分类器决策融合的人脸识别[J].电信科学,2018(4):31-40.
[8]张扬,杨建华,侯宏.一种基于证据理论的数据分聚类融合算法[J].科学技术与工程,2018(1):54-58.
[9]王江,孙美凤,张炜.基于行为知识空间的多分类器网络流量分类方法[J].扬州大学学报:自然科学版,2016(4):54-57.
[10]袁立,穆志纯.基于子分类器融合的部分遮挡人耳识别[J].仪器仪表学报,2011,32(1):186-193.
[11]朱波,陈科,徐君,等.平均分布集成策略:一种新的分类器融合方法[J].小型微型计算机系统,2016,(7):1546-1550.
[12]Borja S P,et al.Ensemble feature selection for rankings of features[J].Springer International Publishing,2015,9095:29-42.
[13]陶晓玲,亢蕊楠,刘丽燕.基于选择性集成的并行多分类器融合方法[J].计算机工程与科学,2018(5):787-792.
[14]刘明,袁保宗,苗振江,等.从局部分类精度到分类置信度的变换[J].计算机研究与发展,2008,45(9):1612-1619.