Considerable research is now focusing on abrupt Holocene changes at millennial- and centennial-scales all over the world. This study shows that the changes in theHolocene summer monsoon can be divided into two main pe...Considerable research is now focusing on abrupt Holocene changes at millennial- and centennial-scales all over the world. This study shows that the changes in theHolocene summer monsoon can be divided into two main periods, based on climatic proxy records of lake sediments from the Sanjiaocheng section located at the NW boundary of the summer monsoon, Gansu Province. The early and middle Holocene was humid with stronger summer monsoon, while the late Holocene is dry with weak summer monsoon. Detailed results also show a series of millennial- and centennial-scale changes in the intensity of the summer monsoon, with a periodicity of 1600 years during the whole Holocene, and with a periodicity of 800 years during the early Holocene. Ten dry events during the Holocene are associated with the weakening of the summer monsoon. These rapid climatic changes may be representative of a global climatic change pattern during the Holocene.展开更多
Foreword Note:Some of the document contents may involve certain patents,the identification of which is not the responsibility of the institution that releases the document.Main drafting organizations:Tsinghua Universi...Foreword Note:Some of the document contents may involve certain patents,the identification of which is not the responsibility of the institution that releases the document.Main drafting organizations:Tsinghua University,Specialty Committee of Network Pharmacology of World Federation of Chinese Medicine Societies(WFCMS).展开更多
文本分类是信息检索的关键问题之一.提取更多的可信反例和构造准确高效的分类器是PU(positive and unlabeled)文本分类的两个重要问题.然而,在现有的可信反例提取方法中,很多方法提取的可信反例数量较少,构建的分类器质量有待提高.分别...文本分类是信息检索的关键问题之一.提取更多的可信反例和构造准确高效的分类器是PU(positive and unlabeled)文本分类的两个重要问题.然而,在现有的可信反例提取方法中,很多方法提取的可信反例数量较少,构建的分类器质量有待提高.分别针对这两个重要步骤提供了一种基于聚类的半监督主动分类方法.与传统的反例提取方法不同,利用聚类技术和正例文档应与反例文档共享尽可能少的特征项这一特点,从未标识数据集中尽可能多地移除正例,从而可以获得更多的可信反例.结合SVM主动学习和改进的Rocchio构建分类器,并采用改进的TFIDF(term frequency inverse document frequency)进行特征提取,可以显著提高分类的准确度.分别在3个不同的数据集中测试了分类结果(RCV1,Reuters-21578,20 Newsgoups).实验结果表明,基于聚类寻找可信反例可以在保持较低错误率的情况下获取更多的可信反例,而且主动学习方法的引入也显著提升了分类精度.展开更多
基金This study was supported by the National Natural Science Foundation of China (Grant No. 49731010)the National Key Basic Research Project on Western Chinese Arid Areas (Grant No. G1999043501).
文摘Considerable research is now focusing on abrupt Holocene changes at millennial- and centennial-scales all over the world. This study shows that the changes in theHolocene summer monsoon can be divided into two main periods, based on climatic proxy records of lake sediments from the Sanjiaocheng section located at the NW boundary of the summer monsoon, Gansu Province. The early and middle Holocene was humid with stronger summer monsoon, while the late Holocene is dry with weak summer monsoon. Detailed results also show a series of millennial- and centennial-scale changes in the intensity of the summer monsoon, with a periodicity of 1600 years during the whole Holocene, and with a periodicity of 800 years during the early Holocene. Ten dry events during the Holocene are associated with the weakening of the summer monsoon. These rapid climatic changes may be representative of a global climatic change pattern during the Holocene.
文摘Foreword Note:Some of the document contents may involve certain patents,the identification of which is not the responsibility of the institution that releases the document.Main drafting organizations:Tsinghua University,Specialty Committee of Network Pharmacology of World Federation of Chinese Medicine Societies(WFCMS).
文摘文本分类是信息检索的关键问题之一.提取更多的可信反例和构造准确高效的分类器是PU(positive and unlabeled)文本分类的两个重要问题.然而,在现有的可信反例提取方法中,很多方法提取的可信反例数量较少,构建的分类器质量有待提高.分别针对这两个重要步骤提供了一种基于聚类的半监督主动分类方法.与传统的反例提取方法不同,利用聚类技术和正例文档应与反例文档共享尽可能少的特征项这一特点,从未标识数据集中尽可能多地移除正例,从而可以获得更多的可信反例.结合SVM主动学习和改进的Rocchio构建分类器,并采用改进的TFIDF(term frequency inverse document frequency)进行特征提取,可以显著提高分类的准确度.分别在3个不同的数据集中测试了分类结果(RCV1,Reuters-21578,20 Newsgoups).实验结果表明,基于聚类寻找可信反例可以在保持较低错误率的情况下获取更多的可信反例,而且主动学习方法的引入也显著提升了分类精度.
