传统基于向量空间模型的文本分类方法需要对文档进行预处理,同时也会损失很多有用的信息。该文提出一种基于离散核支持向量机的文本分类方法,直接根据文档的字符序列构造离散核,用于支持向量机分类算法,比较文档之间的相似性,从而改善...传统基于向量空间模型的文本分类方法需要对文档进行预处理,同时也会损失很多有用的信息。该文提出一种基于离散核支持向量机的文本分类方法,直接根据文档的字符序列构造离散核,用于支持向量机分类算法,比较文档之间的相似性,从而改善文本分类的效果。证明了离散核支持向量机方法的时间复杂度与文本的长度成O(n)关系。在R eu ters-21578文档集上将离散核方法与多项式核、高斯核方法进行比较,实验结果表明该文所提方法在简化分类方法的同时也可以提高分类的精度。展开更多
For domains composed by balls in C^n, this paper studies the boundary behaviour of Cauchy type integrals with discrete holomorphic kernels and the corresponding linear singular integral equation on each piece of smoot...For domains composed by balls in C^n, this paper studies the boundary behaviour of Cauchy type integrals with discrete holomorphic kernels and the corresponding linear singular integral equation on each piece of smooth lower dimensional edges on the boundary of the domain.展开更多
Hydraulic fracturing （HF） technique has been extensively used for the exploitation of unconventional oiland gas reservoirs. HF enhances the connectivity of less permeable oil and gas-bearing rock formationsby fluid ...Hydraulic fracturing （HF） technique has been extensively used for the exploitation of unconventional oiland gas reservoirs. HF enhances the connectivity of less permeable oil and gas-bearing rock formationsby fluid injection, which creates an interconnected fracture network and increases the hydrocarbonproduction. Meanwhile, microseismic （MS） monitoring is one of the most effective approaches to evaluatesuch stimulation process. In this paper, the combined finite-discrete element method （FDEM） isadopted to numerically simulate HF and associated MS. Several post-processing tools, includingfrequency-magnitude distribution （b-value）, fractal dimension （D-value）, and seismic events clustering,are utilized to interpret numerical results. A non-parametric clustering algorithm designed specificallyfor FDEM is used to reduce the mesh dependency and extract more realistic seismic information.Simulation results indicated that at the local scale, the HF process tends to propagate following the rockmass discontinuities; while at the reservoir scale, it tends to develop in the direction parallel to themaximum in-situ stress. 2014 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting byElsevier B.V. All rights reserved.展开更多
文摘传统基于向量空间模型的文本分类方法需要对文档进行预处理,同时也会损失很多有用的信息。该文提出一种基于离散核支持向量机的文本分类方法,直接根据文档的字符序列构造离散核,用于支持向量机分类算法,比较文档之间的相似性,从而改善文本分类的效果。证明了离散核支持向量机方法的时间复杂度与文本的长度成O(n)关系。在R eu ters-21578文档集上将离散核方法与多项式核、高斯核方法进行比较,实验结果表明该文所提方法在简化分类方法的同时也可以提高分类的精度。
基金Project supported by the National Science Foundation of China (10271097)
文摘For domains composed by balls in C^n, this paper studies the boundary behaviour of Cauchy type integrals with discrete holomorphic kernels and the corresponding linear singular integral equation on each piece of smooth lower dimensional edges on the boundary of the domain.
基金supported by the Natural Sciences and Engineering Research Council of Canada through Discovery Grant 341275 (G. Grasselli) and Engage EGP 461019-13
文摘Hydraulic fracturing （HF） technique has been extensively used for the exploitation of unconventional oiland gas reservoirs. HF enhances the connectivity of less permeable oil and gas-bearing rock formationsby fluid injection, which creates an interconnected fracture network and increases the hydrocarbonproduction. Meanwhile, microseismic （MS） monitoring is one of the most effective approaches to evaluatesuch stimulation process. In this paper, the combined finite-discrete element method （FDEM） isadopted to numerically simulate HF and associated MS. Several post-processing tools, includingfrequency-magnitude distribution （b-value）, fractal dimension （D-value）, and seismic events clustering,are utilized to interpret numerical results. A non-parametric clustering algorithm designed specificallyfor FDEM is used to reduce the mesh dependency and extract more realistic seismic information.Simulation results indicated that at the local scale, the HF process tends to propagate following the rockmass discontinuities; while at the reservoir scale, it tends to develop in the direction parallel to themaximum in-situ stress. 2014 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting byElsevier B.V. All rights reserved.
文摘针对多通道四类运动想象(Motor imagery,MI)脑电信号(Electroencephalography,EEG)的分类问题,提出免疫多域特征融合的多核学习SVM(Support vector machine)运动想象脑电信号分类算法.首先,通过离散小波变换(Discrete wavelet transform,DWT)提取脑电信号的时频域特征,并利用一对多公共空间模式(One versus the rest common spatial patterns,OVR-CSP)提取脑电信号的空域特征,融合时频空域特征形成特征向量.其次,利用多核学习支持向量机(Multiple kernel learning support vector machine,MKL-SVM)对提取的特征向量进行分类.最后,利用免疫遗传算法(Immune genetic algorithm,IGA)对模型的相关参数进行优化,得到识别率更高的脑电信号分类模型.采用BCI2005desc-Ⅲa数据集进行实验验证,对比结果表明,本文所提出的分类模型有效地解决了传统单域特征提取算法特征单一、信息描述不足的问题,更准确地表达了不同受试者个性化的多域特征,取得了94.21%的识别率,优于使用相同数据集的其他方法.
