Communication optimization is very important for imporoving performance of parallel programs A communication optimization method called HVMP(Half Vector Message Ripelining) is presented. In comparison with the widelyu...Communication optimization is very important for imporoving performance of parallel programs A communication optimization method called HVMP(Half Vector Message Ripelining) is presented. In comparison with the widelyused vector message pipelining, HVMP can get better tradeoff between reducing and hiding communication overhead,and eliminate the communication barrier of barrier synchronization problems[1]. For parallel Systems with low bandwidth such as cluster of workstations and barrier synchronization problems with large amount of communication, HVMPmethod can get good performance.展开更多
准确的地表覆盖信息是陆地过程模拟与生态系统管理的重要基础。近年来多种地表覆盖产品相继发布,但因分类体系、数据源与算法差异,在区域尺度应用中存在较大不确定性,而现有研究多关注产品间的精度差异,较少探讨如何整合多源产品优势以...准确的地表覆盖信息是陆地过程模拟与生态系统管理的重要基础。近年来多种地表覆盖产品相继发布,但因分类体系、数据源与算法差异,在区域尺度应用中存在较大不确定性,而现有研究多关注产品间的精度差异,较少探讨如何整合多源产品优势以构建高精度融合产品。本文以气候变化倡议(climate change initiative,CCI)、中国土地覆盖数据集(China land cover dataset,CLCD)、全球地表覆盖30 m分辨率精细分类系统(global land cover with fine classification system at 30 m resolution,GLC_FCS30)和MCD12Q1四套地表覆盖产品为基础,设计一种兼顾一致性与精度的多源地表覆盖产品融合方法。通过提取各产品一致区域作为共识区;随后结合F1分数对不一致区域进行加权判别,优选最可信类别,生成精度更高的融合产品;为验证方法可行性,在川滇生态屏障区进行验证。结果显示:不同产品具有较高的一致性,但分类精度差异明显;融合产品的空间分布与原始产品相似,但整体分类精度更高,总体精度提高了约8.04%,达到81.34%,Kappa系数提高了约14.13%,达到69.1%,并在森林、草地等关键地类上表现更优,有效降低了类别间的混淆程度。展开更多
文摘Communication optimization is very important for imporoving performance of parallel programs A communication optimization method called HVMP(Half Vector Message Ripelining) is presented. In comparison with the widelyused vector message pipelining, HVMP can get better tradeoff between reducing and hiding communication overhead,and eliminate the communication barrier of barrier synchronization problems[1]. For parallel Systems with low bandwidth such as cluster of workstations and barrier synchronization problems with large amount of communication, HVMPmethod can get good performance.
文摘准确的地表覆盖信息是陆地过程模拟与生态系统管理的重要基础。近年来多种地表覆盖产品相继发布,但因分类体系、数据源与算法差异,在区域尺度应用中存在较大不确定性,而现有研究多关注产品间的精度差异,较少探讨如何整合多源产品优势以构建高精度融合产品。本文以气候变化倡议(climate change initiative,CCI)、中国土地覆盖数据集(China land cover dataset,CLCD)、全球地表覆盖30 m分辨率精细分类系统(global land cover with fine classification system at 30 m resolution,GLC_FCS30)和MCD12Q1四套地表覆盖产品为基础,设计一种兼顾一致性与精度的多源地表覆盖产品融合方法。通过提取各产品一致区域作为共识区;随后结合F1分数对不一致区域进行加权判别,优选最可信类别,生成精度更高的融合产品;为验证方法可行性,在川滇生态屏障区进行验证。结果显示:不同产品具有较高的一致性,但分类精度差异明显;融合产品的空间分布与原始产品相似,但整体分类精度更高,总体精度提高了约8.04%,达到81.34%,Kappa系数提高了约14.13%,达到69.1%,并在森林、草地等关键地类上表现更优,有效降低了类别间的混淆程度。
