二维码（扫一下试试看！） | | 融合子集特征级联预学习的封装方法研究 | | A Wrapper Method for Combining Subset Feature with Cascade Pre-learning | | 投稿时间：2019-04-13 | | DOI：10.15918/j.tbit1001-0645.2019.117 | | 中文关键词:特征选择封装法级联森林特征学习 | | English Keywords:feature selectionwrapper methodcascade forestfeature learning | | 基金项目:国家"十三五"科技支撑计划项目（SQ2018YFC200004）；国家卫生部卫生行业科研专项基金项目（201302008） | | | 摘要点击次数:67 | | 全文下载次数:51 | | 中文摘要: | | 机器学习领域中的特征选择算法可简化模型输入，提高可解释性并帮助避免维度灾难及过拟合现象的发生.针对基于封装法进行特征选择时，评价模型通常将搜索出的特征子集直接作为输入，导致算法对特征利用和评估效果受限于评价模型的特征学习能力，限制了对更适特征子集的发现能力等问题，提出一种基于级联森林结构的子集特征预学习封装法.该方法在搜索算法与评价模型之间添加多层级联森林，重构待评价特征子集为高级特征集，降低评价模型模式识别难度，提高对子集性能的评价效果.实验对比了多种搜索算法及评价模型组合，本方法可在保证分类性能的前提下，进一步降低所选特征数量，同时维持了封装法的低耦合性. | | English Summary: | | Feature selection algorithms in the machine learning domain can simplify the input of model, improve interpretability, and avoid dimensional catastrophe and over-fitting. In terms of selecting features on wrapper methods, the evaluation of models usually take the feature subsets filtered by the search algorithm as input directly, which leads to the fact that feature exploitation and evaluation of models is restricted by the feature reconstruction and fitting ability of the evaluation model. Moreover, the more appropriate feature subsets were limited to be discovered either. To solve the problems, a pre-learning wrapper method was proposed based on the cascade forest structure. Adding multi-level cascade forest between the search algorithm and the evaluation model, the model was arranged to transform the feature subset as high-level feature set, reducing the difficulty of recognition in the evaluation and improving the performance of feature subset. In contrast experiment, a variety of search algorithms and evaluation model combinations were evaluated on multiple datasets. The results indicate that the proposed method can reduce the number of selected features, while maintaining classification performance and the low coupling property of wrapper methods. | | 查看全文查看/发表评论下载PDF阅读器 | |
高原,施云惠,韩妍妍,曾萍,尹宝才.附加法向信息的三维网格预测编码[J].北京理工大学学报（自然科学版）,2019,39(1):88~94.GAOYuan,SHIYun-hui,HANYan-yan,ZENGPing,YINBao-cai.Compressionof3DMeshBasedonNorm ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21钟巍,寿列枫,王仲琦,田宙,刘俊,何增,李伟昌.PVB夹层钢化玻璃冲击波毁伤效应实验研究[J].北京理工大学学报（自然科学版）,2019,39(6):565~570.ZHONGWei,SHOULie-feng,WANGZhong-qi,TIANZhou,LIUJun,HEZeng,LIWei-cha ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21张冬晓,陈亚洲,程二威,杜宝舟.无人机信息链路电磁干扰效应规律研究[J].北京理工大学学报（自然科学版）,2019,39(7):756~762.ZHANGDong-xiao,CHENYa-zhou,CHENGEr-wei,DUBao-zhou.EffectsofElectromagneticInte ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21由育阳,由书凯,高健凯,杨志宏.基于正态逆高斯和特征贡献度的睡眠分期实验研究[J].北京理工大学学报（自然科学版）,2019,39(8):833~838.YOUYu-yang,YOUShu-kai,GAOJian-kai,YANGZhi-hong.ExperimentalStudyonSleepSt ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21任会兰,安刚,郝莉,王宗炼.铝合金PLC效应与声发射特性的实验研究[J].北京理工大学学报（自然科学版）,2019,39(10):999~1005.RENHui-lan,ANGang,HAOLi,WANGZong-lian.ExperimentalStudyonPLCEffectandAcousti ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21李梅,马宇宇,蒋建伟,郗莉萍.双爆源气泡与水面相互作用的实验[J].北京理工大学学报（自然科学版）,2019,39(12):1219~1224.LIMei,MAYu-yu,JIANGJian-wei,XILi-ping.ExperimentsoftheInteractionsBetweenDoubl ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21汪亦显,汪志强,张伟,刘飞飞.含预制孔洞混凝土的断裂实验研究[J].北京理工大学学报（自然科学版）,2019,39(12):1232~1238.WANGYi-xian,WANGZhi-qiang,ZHANGWei,LIUFei-fei.ExperimentalResearchonFractureof ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21.北京理工大学学报2019年总目次（第39卷）[J].北京理工大学学报（自然科学版）,2019,39(12):1321~1338..[J].TransactionsofBeijingInstituteofTechnology,2019,39(12):1321-1338.二维码（扫一下试试看！）北京理 ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21王磊,张天恒,郭苗苗,张颖,杨硕,徐桂芝.声音骨传导方式与空气传导方式在听觉诱发脑电实验中的对比[J].北京理工大学学报（自然科学版）,2019,39(S1):28~32.WANGLei,ZHANGTian-heng,GUOMiao-miao,ZHANGYing,YANGShuo,XUGui-zhi ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21李海松,蒋轶虎,杨博,岳红菊,唐威.CMOS/SOI工艺触发器单元的单粒子实验验证与分析[J].北京理工大学学报（自然科学版）,2018,38(1):63~67.LIHai-song,JIANGYi-hu,YANGBo,YUEHong-ju,TANGWei.SEEEvaluationMethodfo ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21
| |
