期刊文献+
共找到3篇文章
< 1 >
每页显示 20 50 100
一种基于ECVM的Tri-training半监督垃圾邮件检测算法
1
作者 卜华龙 夏静 郑尚志 《宿州学院学报》 2016年第8期105-107,117,共4页
为提高垃圾邮件检测精度,提出一种基于ECVM的Tri-training半监督垃圾邮件检测算法,兼顾了Tri-training算法的准确性和ECVM算法处理大规模数据的高效性特点,可以降低算法的时间和空间复杂度,提高未标记数据的利用率,适应垃圾邮件数据的... 为提高垃圾邮件检测精度,提出一种基于ECVM的Tri-training半监督垃圾邮件检测算法,兼顾了Tri-training算法的准确性和ECVM算法处理大规模数据的高效性特点,可以降低算法的时间和空间复杂度,提高未标记数据的利用率,适应垃圾邮件数据的规模大、标记数据少、稀疏性强等特点。Matlab实验表明Tri-training+ECVM比传统的Tri-training+SVM在准确率和时间复杂度指标上都有大幅度的提升。 展开更多
关键词 TRI-TRAINING ecvm 垃圾邮件检测 半监督学习
在线阅读 下载PDF
一种适用于WAMS量测数据的系统暂态功角稳定评估方法 被引量:18
2
作者 邵雅宁 唐飞 +2 位作者 刘涤尘 马志昊 卞成志 《电力系统保护与控制》 EI CSCD 北大核心 2015年第6期33-39,共7页
考虑WAMS量测数据刷新速率快,数据量大的特点,提出了一种适用于WAMS量测数据的暂态功角稳定评估方法。选取初始特征量集并用核主成分分析法对特征集进行降维,过滤冗余特征并降低分类器输入向量的维度。构建训练样本集,计算各样本的初始... 考虑WAMS量测数据刷新速率快,数据量大的特点,提出了一种适用于WAMS量测数据的暂态功角稳定评估方法。选取初始特征量集并用核主成分分析法对特征集进行降维,过滤冗余特征并降低分类器输入向量的维度。构建训练样本集,计算各样本的初始特征量集并进行降维。通过训练ECVM分类器对暂态功角稳定进行评估,并用测试数据集验证分类器的准确率。在新英格兰10机39节点系统中的仿真表明,所提算法有较高的分类准确率,与传统分类算法相比降低了单个样本评估所需的时间,具有工程使用价值。 展开更多
关键词 WAMS数据 核主成分分析 ecvm分类器 暂态稳定评估
在线阅读 下载PDF
Identifying the turbulent flow developing inside and around the bottom trawl by Electromagnetic Current Velocity Meter approach in the flume tank 被引量:4
3
作者 Nyatchouba Nsangue Bruno Thierry Hao Tang +4 位作者 Liu-xiong Xu Fuxiang Hu Xinxing You Micah Adekunle David Njomoue Pandong Achille 《Journal of Hydrodynamics》 SCIE EI CSCD 2021年第3期636-656,共21页
The Knowledge of turbulent flow developing inside and around the bottom trawl net is of great importance not only for improving the hydrodynamic performance of the gear but also for the selectivity via the fish respon... The Knowledge of turbulent flow developing inside and around the bottom trawl net is of great importance not only for improving the hydrodynamic performance of the gear but also for the selectivity via the fish response,such as the herding response or escape behavior.The 3-D Electromagnetic Current Velocity Meter(ECVM)measurements were performed to investigate the effect of turbulent flow on the bottom trawl net performance and to analyze the turbulence intensity and velocity ratio inside and around different parts of the trawl net.Proper orthogonal decomposition(POD)method was applied in order to extract the phase averaged mean velocity field of turbulent flow from each available ECVM instantaneous velocity.The results demonstrated the existence of turbulence flow,consisting of turbulent boundary layer flow and the turbulence due to the trawl wake developing all inside and around the bottom trawl net.Increasing input streamwise velocity results in faster trawl movement and a significant turbulent flow.The maximum turbulence intensity inside and around trawl wing,square part,first belly,second belly,third belly,cod-end is 0.95%,1.34%,3.40%,4.10%,4.25%and 3.80%,respectively.It was found that the mean velocity field in a turbulent flow inside and around trawl net cod-end recovered on the average was~77.58%of the input streamwise velocity.It is~12.92%,~13.07%,~11.40%,~13.00%and~0.45%less than that inside and around trawl wing,square part,first belly,second belly,and third belly of the bottom trawl net,respectively.The turbulent flow behavior depends strongly on the structure oscillation,input streamwise velocity and,porosity of the net structure.It is necessary to take into account the velocity reduction inside and around a different part of the trawl net to improve the entire drag force determination,cod-end design,and further selectivity control of the fishing gear. 展开更多
关键词 HYDRODYNAMIC bottom trawl net turbulent flow Electromagnetic Current Velocity Meter(ecvm) turbulence intensity proper orthogonal decomposition(POD)
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部