期刊文献+
共找到5篇文章
< 1 >
每页显示 20 50 100
带重现概念漂移的不平衡数据流分类研究 被引量:2
1
作者 季梦遥 袁磊 《贵州大学学报(自然科学版)》 2019年第3期69-74,共6页
数据流广泛存在于现实应用中,重现概念漂移和数据分布不平衡性是其重要特性,它们会导致传统数据流分类器负偏离、性能下降及学习时间倍增。针对重现概念漂移和数据分布不平衡的特点,本文提出重现概念漂移不平衡数据流的随机平衡采样集... 数据流广泛存在于现实应用中,重现概念漂移和数据分布不平衡性是其重要特性,它们会导致传统数据流分类器负偏离、性能下降及学习时间倍增。针对重现概念漂移和数据分布不平衡的特点,本文提出重现概念漂移不平衡数据流的随机平衡采样集成分类算法(RBSRISEA),首先用随机平衡采样算法重新平衡数据分布,之后对预处理的数据流再进行重现概念漂移探测。实验表明,RBSRISEA对重现概念漂移有较强的敏感性和泛化能力。RBSRISEA可以处理带重现概念漂移的不平衡数据流分类问题。 展开更多
关键词 重现概念漂移 不平衡流数据 随机采样 集成分类器
下载PDF
概念漂移不平衡数据流随机平衡采样分类算法
2
作者 袁磊 季梦遥 《湖北大学学报(自然科学版)》 CAS 2019年第1期95-100,共6页
带概念漂移不平衡流数据分类研究是机器学习和现实应用领域的一个难点和热点.针对带概念漂移不平衡流数据的动态性和不平衡性,本文中提出了随机平衡采样算法用于再平衡不平衡数据流.之后,在随机平衡采样算法的基础上提出了一种新的处理... 带概念漂移不平衡流数据分类研究是机器学习和现实应用领域的一个难点和热点.针对带概念漂移不平衡流数据的动态性和不平衡性,本文中提出了随机平衡采样算法用于再平衡不平衡数据流.之后,在随机平衡采样算法的基础上提出了一种新的处理带概念漂移的不平衡流数据集成分类算法用于抵抗流数据的概念漂移和不平衡性.理论和实验表明本文中提出的集成分类算法对处理带概念漂移的不平衡流数据较强的多样性和泛化能力. 展开更多
关键词 概念漂移 不平衡流数据 采样 集成分类器
下载PDF
基于GA启发式抽样的交通事件自动检测 被引量:1
3
作者 李苗华 陈淑燕 +1 位作者 劳叶春 谷健 《交通信息与安全》 2016年第5期87-92,共6页
为了提高面向不平衡数据集的交通事件检测综合性能,提出了两种基于GA启发式抽样方法的交通事件检测算法。基于GA的实例选择抽样方法(GA-IS),解决非启发式抽样方法人为设定抽样率导致的检测效果不稳定问题。基于GA的支持向量选择抽样方法... 为了提高面向不平衡数据集的交通事件检测综合性能,提出了两种基于GA启发式抽样方法的交通事件检测算法。基于GA的实例选择抽样方法(GA-IS),解决非启发式抽样方法人为设定抽样率导致的检测效果不稳定问题。基于GA的支持向量选择抽样方法(GA-SS),改善学习集数据量较大时的检测效率。实验采用新加坡AYE仿真数据库,以支持向量机作为分类器进行事件检测。结果表明,基于遗传算法实例选择抽样的检测模型检测率达到94%,平均检测时间为1.413 3min,性能指标PI为0.157;基于遗传算法支持向量选择抽样的检测模型决策时间为4.55s,综合性能最优,其PI为0.151;基于少数类过抽样算法(SMOTE)的检测模型决策时间为35.21s,PI为0.329,与非启发式抽样方法相比,所提方法能有效改善面向不平衡数据集的事件检测综合性能。 展开更多
关键词 交通安全 交通事件检测 不平交通数据 启发式抽样 遗传算法
下载PDF
Energy-balanced clustering protocol for data gathering in wireless sensor networks with unbalanced traffic load 被引量:1
4
作者 奎晓燕 王建新 张士庚 《Journal of Central South University》 SCIE EI CAS 2012年第11期3180-3187,共8页
Energy-efficient data gathering in multi-hop wireless sensor networks was studied,considering that different node produces different amounts of data in realistic environments.A novel dominating set based clustering pr... Energy-efficient data gathering in multi-hop wireless sensor networks was studied,considering that different node produces different amounts of data in realistic environments.A novel dominating set based clustering protocol (DSCP) was proposed to solve the data gathering problem in this scenario.In DSCP,a node evaluates the potential lifetime of the network (from its local point of view) assuming that it acts as the cluster head,and claims to be a tentative cluster head if it maximizes the potential lifetime.When evaluating the potential lifetime of the network,a node considers not only its remaining energy,but also other factors including its traffic load,the number of its neighbors,and the traffic loads of its neighbors.A tentative cluster head becomes a final cluster head with a probability inversely proportional to the number of tentative cluster heads that cover its neighbors.The protocol can terminate in O(n/lg n) steps,and its total message complexity is O(n2/lg n).Simulation results show that DSCP can effectively prolong the lifetime of the network in multi-hop networks with unbalanced traffic load.Compared with EECT,the network lifetime is prolonged by 56.6% in average. 展开更多
关键词 ENERGY-BALANCE CLUSTERING data gathering wireless sensor networks unbalanced traffic load
下载PDF
On the atmospheric movement and the imbalance of observed and calculated energy in the surface layer 被引量:5
5
作者 ZUO HongChao XIAO Xia +3 位作者 YANG QiDong DONG LongXiang CHEN JiWei WANG ShuJin 《Science China Earth Sciences》 SCIE EI CAS 2012年第9期1518-1532,共15页
Based on existing researches,here we theoretically summarized the characteristics of the atmospheric movement and turbulent transport of energy and substance in the surface layer as well as the ideal and the actual mo... Based on existing researches,here we theoretically summarized the characteristics of the atmospheric movement and turbulent transport of energy and substance in the surface layer as well as the ideal and the actual models for the turbulent transport.Then,using the data observed with eddy covariance at the semiarid climate and environment monitoring station(SACOL) in Lanzhou University from May to October during four consecutive years(September 2006-August 2010),we conducted a detailed analysis of the turbulent transport in the surface layer,through introducing the relative vertical turbulence intensity to characterize the turbulence strength,RIw=wn(wn+U),and also by adopting the method for controlling data quality at different levels.Our conclusions are:(1) The turbulent transport of energy and substance in the surface layer must obey the law of conservation of energy and the law of conservation of matter,the observed and calculated energy in the surface layer must be balanced,or closed in theory,but the actual observed and calculated energy just approximates the ideal in some degree and is difficult to achieve the energy balance.(2) The energy closure rate depends much on the atmospheric state in the surface layer,and the energy closure rate increases generally with the relative vertical turbulence intensity.(3) By the way of controlling data quality at different levels,it is found that the degree of data quality control can affect the closure rate,but it does not change the fact that the energy closure rate depends on the atmospheric state.(4) The calculation method of surface soil heat flux can affect energy closure rate,but does not change its dependence on the atmospheric state. 展开更多
关键词 surface layer energy balance state relative vertical turbulence intensity turbulence transport data quality control
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部