针对非平行支持向量机(NonParallel Support Vector Machine,NPSVM)对噪声敏感和忽略了数据分布结构的问题,提出了一种具有间隔分布的抗噪声非平行支持向量机(Anti-Noise NPSVM with Margin Distribution, MDANPSVM)分类模型.在MD-ANPSV...针对非平行支持向量机(NonParallel Support Vector Machine,NPSVM)对噪声敏感和忽略了数据分布结构的问题,提出了一种具有间隔分布的抗噪声非平行支持向量机(Anti-Noise NPSVM with Margin Distribution, MDANPSVM)分类模型.在MD-ANPSVM模型中,每个优化问题同时最小化两类样本的基于L1范数的绝对损失和改进的铰链损失,这可以保证模型的稳定性,减小噪声和异常值的影响.此外,在MD-ANPSVM模型中,采用一阶和二阶统计量来描述训练数据的间隔分布信息,并试图同时最大化间隔均值和最小化间隔方差,这进一步提高了模型的泛化性能.最终,我们在不同的数据集上进行了对比实验.实验结果显示,MD-ANPSVM模型具有较强的泛化能力和强鲁棒性.展开更多
Formed on top of the Gulf of Cadiz, the Al Idrissi mud volcano is the shallowest and largest mud volcano in the El Arraiche mud volcano field of the northwestern Moroccan margin. The development and morphology of mud ...Formed on top of the Gulf of Cadiz, the Al Idrissi mud volcano is the shallowest and largest mud volcano in the El Arraiche mud volcano field of the northwestern Moroccan margin. The development and morphology of mud volcanoes from the El Arraiche mud volcanoes group have been studied at a large scale. However, the time interval related to their formation period still needs to be better understood. In this regard, we interpreted and analyzed the seismic facies from the 2D reflection data of the GEOMARGEN-1 campaign, which took place in 2011. The aim was to identify the seismic sequences and draw the Al Idrissi mud volcano system to determine the formation period of the Al Idriss mud volcano. And as a result, the Al Idrissi mud volcano system is made of both buried and superficial bicone and was identified along with the Upper Tortonian to Messinian-Upper Pliocene facies. As the initial mud volcano extrusive edifice, the buried bicone was formed in the Late-Messinian to Early-Pliocene period. However, the superficial bicone, as the final extrusive edifice, was included in the Late Pliocene. In this case, the timing interval between the buried and superficial bicone is equivalent to the Late-Messinian to Upper-Pliocene period. Therefore, the latter corresponds to the Al Idrissi mud volcano formation period.展开更多
Substance P is an endogenous neurokinin that is present in the central and peripheral nervous systems. The neuropeptide substance P and its high-affinity receptor neurokinin 1 receptor are known to play an important r...Substance P is an endogenous neurokinin that is present in the central and peripheral nervous systems. The neuropeptide substance P and its high-affinity receptor neurokinin 1 receptor are known to play an important role in the central nervous system in inflammation, blood pressure, motor behavior and anxiety. The effects of substance P in the hippocampus and the marginal di- vision of the striatum on memory remain poorly understood. Compared with the hippocampus as a control, immunofluorescence showed high expression of the substance P receptor, neuro- kinin 1, in the marginal division of the striatum of normal rats. Unilateral or bilateral injection of an antisense oligonucleotide against neurokinin 1 receptor mRNA in the rat hippocampus or marginal division of the striatum effectively reduced neurokinin 1 receptor expression. Indepen- dent of injection site, rats that received this antisense oligonucleotide showed obviously increased footshock times in a Y-maze test. These results indicate that the marginal division of the striatum plays a similar function in learning and memory to the hippocampus, which is a valuable addi- tion to our mechanistic understanding of the learning and memory functions of the marginal division of the striatum.展开更多
针对高光谱数据波段多,地物标签获取代价高,带标记的样本数量少,分类过程中容易引起Hudges现象。本文提出一种基于改进的局部全局一致性(learning with local and global consistency,LLGC)算法的半监督分类方法。通过边缘采样法(margin...针对高光谱数据波段多,地物标签获取代价高,带标记的样本数量少,分类过程中容易引起Hudges现象。本文提出一种基于改进的局部全局一致性(learning with local and global consistency,LLGC)算法的半监督分类方法。通过边缘采样法(margin sampling,MS)选取最富含信息量的无标签样本,加入到训练集来扩充训练样本;用KNN算法计算相似度进一步优选无标签样本,去除噪声点和存在的野值点;使用改进的局部全局一致性算法对无标签样本集进行分类标记,得到各类别的分类结果。实验结果表明,本文方法在充分利用无标签样本的情况下,有效地提高了带有少量标签样本的高光谱图像的分类精度。展开更多
文摘针对非平行支持向量机(NonParallel Support Vector Machine,NPSVM)对噪声敏感和忽略了数据分布结构的问题,提出了一种具有间隔分布的抗噪声非平行支持向量机(Anti-Noise NPSVM with Margin Distribution, MDANPSVM)分类模型.在MD-ANPSVM模型中,每个优化问题同时最小化两类样本的基于L1范数的绝对损失和改进的铰链损失,这可以保证模型的稳定性,减小噪声和异常值的影响.此外,在MD-ANPSVM模型中,采用一阶和二阶统计量来描述训练数据的间隔分布信息,并试图同时最大化间隔均值和最小化间隔方差,这进一步提高了模型的泛化性能.最终,我们在不同的数据集上进行了对比实验.实验结果显示,MD-ANPSVM模型具有较强的泛化能力和强鲁棒性.
文摘Formed on top of the Gulf of Cadiz, the Al Idrissi mud volcano is the shallowest and largest mud volcano in the El Arraiche mud volcano field of the northwestern Moroccan margin. The development and morphology of mud volcanoes from the El Arraiche mud volcanoes group have been studied at a large scale. However, the time interval related to their formation period still needs to be better understood. In this regard, we interpreted and analyzed the seismic facies from the 2D reflection data of the GEOMARGEN-1 campaign, which took place in 2011. The aim was to identify the seismic sequences and draw the Al Idrissi mud volcano system to determine the formation period of the Al Idriss mud volcano. And as a result, the Al Idrissi mud volcano system is made of both buried and superficial bicone and was identified along with the Upper Tortonian to Messinian-Upper Pliocene facies. As the initial mud volcano extrusive edifice, the buried bicone was formed in the Late-Messinian to Early-Pliocene period. However, the superficial bicone, as the final extrusive edifice, was included in the Late Pliocene. In this case, the timing interval between the buried and superficial bicone is equivalent to the Late-Messinian to Upper-Pliocene period. Therefore, the latter corresponds to the Al Idrissi mud volcano formation period.
基金supported by the National Natural Science Foundation of China,No.30600797,30873238
文摘Substance P is an endogenous neurokinin that is present in the central and peripheral nervous systems. The neuropeptide substance P and its high-affinity receptor neurokinin 1 receptor are known to play an important role in the central nervous system in inflammation, blood pressure, motor behavior and anxiety. The effects of substance P in the hippocampus and the marginal di- vision of the striatum on memory remain poorly understood. Compared with the hippocampus as a control, immunofluorescence showed high expression of the substance P receptor, neuro- kinin 1, in the marginal division of the striatum of normal rats. Unilateral or bilateral injection of an antisense oligonucleotide against neurokinin 1 receptor mRNA in the rat hippocampus or marginal division of the striatum effectively reduced neurokinin 1 receptor expression. Indepen- dent of injection site, rats that received this antisense oligonucleotide showed obviously increased footshock times in a Y-maze test. These results indicate that the marginal division of the striatum plays a similar function in learning and memory to the hippocampus, which is a valuable addi- tion to our mechanistic understanding of the learning and memory functions of the marginal division of the striatum.
文摘针对高光谱数据波段多,地物标签获取代价高,带标记的样本数量少,分类过程中容易引起Hudges现象。本文提出一种基于改进的局部全局一致性(learning with local and global consistency,LLGC)算法的半监督分类方法。通过边缘采样法(margin sampling,MS)选取最富含信息量的无标签样本,加入到训练集来扩充训练样本;用KNN算法计算相似度进一步优选无标签样本,去除噪声点和存在的野值点;使用改进的局部全局一致性算法对无标签样本集进行分类标记,得到各类别的分类结果。实验结果表明,本文方法在充分利用无标签样本的情况下,有效地提高了带有少量标签样本的高光谱图像的分类精度。
