Seizure detection using earth movers' distance and SVM in intracranial EEG

Seizure detection is extremely essential for long-term monitoring of epileptic patients. This paper investigates the detection of epileptic seizures in multi-channel long-term intracranial electroencephalogram （iEEG）. The algorithm conducts wavelet decomposition of iEEGs with five scales, and transforms the sum of the three frequency bands into histogram for computing the distance. The proposed method combines a novel feature called EMD-L1, which is an efficient algorithm of earth movers＇ distance （EMD）, with support vector machine （SVM） for binary classification between seizures and non-sei- zures. The EMD-LI used in this method is characterized by low time complexity and high processing speed by exploiting the L~ metric structure. The smoothing and collar technique are applied on the raw outputs of SVM classifier to obtain more ac- curate results. Several evaluation criteria are recommended to compare our algorithm with other conventional methods using the same dataset from the Freiburg EEG database. Experiment results show that the proposed method achieves a high sensi- tivity, specificity and low false detection rate, which are 95.73 %, 98.45 % and 0.33/h, respectively. This algorithm is char- acterized by its robustness and high accuracy with the possibility of performing real-time analysis of EEG data, and may serve as a seizure detection tool for monitoring long-term EEG.

Aristotle＇s Immovable Movers： A Sketch

In keeping with a view that is explicitly formulated by Aristotle in his Motion of Animals, general kinetic principles must be specified according to the different types of movable entities existing in the universe. At issue, essentially, are the motions of the stars and the motions of animals. Whereas the cosmological immovable mover is the object of two complementary analyses （in Bk. Ⅷ of Physics and in Chs. 6 and 7 of Bk. Ⅻ of Metaphysics）, information on the immovability of the first mover responsible for animal motion is to be found in the psychological and psycho-physiological treatises （On the Soul, in Bk. Ⅰ, Chs. 3 and 4, and in Bk. Ⅲ, Ch. 10 and in Ch. 6 of the Motion of Animals）. But it is also found in Ch. 7, Bk. Ⅻ of the Metaphysics, in the very context of the argument concerning the absolutely first immovable mover of the world. This suggests that the two types of motion, that of the stars and that of animals, however distinct the arguments about them are, rest on a single scheme, and maybe even on a common principle. This is liable to surprise us, as much as stars and animals appear to us to belong to heterogeneous orders of reality. But the situation is different for Aristotle, who, as attentive as he is to differences, tends nonetheless to conceive the stars as living things of a particular kind. This fact is the source of a series of difficulties that Aristotle generously left for his many commentators to solve. Aim of this text, which was initially directed to a larger audience, is to set some of these complex issues in both simple and up to date terms.

Movers and Shapers

In an exhibition area of the World Economic Forum's (WEF) Annual Meeti ng of the New Cham pions in China's northeastern coastal city of Dalian, a robot was performing a surgical operation on a rose. The volunteers in the booth invited visitors to have a go at running the machine, which was developed by U.S. company Intuitive Surgical and has been used in 6 million successful surgeries worldwide. Over 110,000 of these were done in China, which has around 80 such robots in use in hospitals.

定量方法对比研究一致性评价Bland-Altman法LoA的可信区间估计

目的阐明定量方法对比研究中一致性评价Bland-Altman法进行LoA可信区间估计的必要性,探讨LoA可信区间估计的方法学性能。方法针对采用两种定量测量方法同时测量一份样本、每方法测量一次的情形,假定两种方法测量结果的差值服从标准正态分布,借助Monte-Carlo模拟,在不同的样本量下分别采用Bland-Altman法(简称B-A法)和MOVER法模拟1万次,获得95%LoA范围95%可信区间覆盖-1.96到1.96范围的比例(覆盖率),用以评价不同方法的性能,同时在临床可接受误差限值为-2.5到2.5范围的设定下判定用不同的可信区间方法达到一致的比例(达标率)加以佐证。结果直接采用LoA法评价一致性因没有考虑抽样误差影响,理论上不严谨、实用时太宽松。用B-A法估计LoA可信区间覆盖率具有样本量依赖性,而MOVER法覆盖率独立于样本量、始终紧紧围绕事先设定的可信度95%微小波动。结论直接用LoA判定定量方法对比研究中的一致性并不合适,采用MOVER法进行LoA可信区间估计用于评价一致性具有很好的可靠性,是一种值得推荐的方法。

一种基于部件空间分布的三维模型检索方法

三维形状分析是三维模型检索的关键问题.提出一种基于三维模型部件空间分布的形状特征描述方法.此方法的主要思想是依据认知心理学的理论,在描述对象形状时强调它的结构属性.首先将三维模型分割为若干个组成部件,每个部件用一个曲面片表示,然后采用曲面片的质心位置、面积占总面积的百分比的组合作为部件特征,最后将满足指定条件的部件特征的集合作为三维模型的形状特征.基于这一特征表示,给出了一种三维模型检索方法.该方法具有受模型精度和连通性影响较小、相似性度量的计算速度较快的优点.实验结果验证了该检索方法的有效性.

定量方法对比研究重复测量设计Bland-Altman一致限LoA的可信区间估计

目的本文将介绍两种不同重复测量设计情形一致性评价的Bland-Altman一致限LoA及其可信区间估计方法,并探讨两种不同设计情形中分别涉及的两种不同可信区间估计方法的统计学性能,为方法学的选择应用提供理论依据。方法根据所测量样本真值是否变化,将重复测量设计分为两种类型,分别给出其Bland-Altman一致限LoA及其可信区间估计的两种方法(一种是Bland-Altman法,简称B-A法;另一种是方差估计反推法,简称MOVER法),借助Monte-Carlo模拟技术,在正态分布假定下,设定不同的样本量、不同的个体内变异性,分别采用B-A法和MOVER法,模拟获得95%LoA范围的95%可信区间对总体设定的覆盖率,用覆盖率与95%的接近程度评价不同方法的统计学性能,理论上该覆盖率与95%越接近,反映方法的统计学性能越好。结果用B-A法估计LoA可信区间的覆盖率随样本量的增大而增大,样本量为20以下时覆盖率不足90%,即使在样本量达到250时,其覆盖率仍在94%以下,而MOVER法覆盖率始终紧紧围绕事先设定的可信度95%微小波动,甚至在样本量为10时覆盖率也能接近95%。结论 MOVER法用于方法对比研究重复测量设计的LoA可信区间估计具有很好的统计学性能。

一种基于EMD的文档语义相似性度量

针对基于EMD(Earth Mover's Distance)的文档语义相似性算法不满足度量公理因而难以在信息检索与数据挖掘中推广应用的问题,该文提出了一种新的基于EMD的文档语义相似性度量——..Mdss_EMD(Metric for document semantic similarity based EMD)。首先在分析EMD及现有改进方法缺陷的基础上,给出了文档宽度、虚拟项的概念;随后通过增加虚拟项来对齐文档矢量的总权值,使所有度量公理得到满足;最后,为提高该度量的适应能力及处理速度,还实现了虚拟项相似距离的弹性设计并对EMD算法进行了简化。该方法把EMD扩展到度量空间中来,很大程度上提高了EMD的索引能力与精度,初步实验表明,Mdss_EMD的整体性能优于原EMD及现有其它类似方法。

配对二项资料两组率差置信区间估计方法与评价

目的：探讨配对二项资料两组率差置信区间的估计方法,并从中进行优选推荐。方法：按照方差估计反推（method of variance of estimates recovery,MOVER）原理,将两配对组率的关联Ф系数与单组率的置信区间进行组合,构建两组率差的置信区间估计方法。其中,单组率置信区间估计分别采用Wilson计分法、Agresti-Coull法（AC法）、Jeffreys法和Clopper-Pearson精确法（CP法）。借助Monte Carlo模拟实验比较不同方法的统计学性能,在不同参数设定下进行I类错误率和把握度的模拟实验：（1）设定两组率关联Ф系数为0、0.2、0.4、0.6,样本量为20、60、100,分别模拟不同率水平下各方法的I类错误率,判定其模拟I类错误率是否接近事先定义的检验水平。（2）设定关联Ф系数为0.3,两组率差为10%,分别模拟不同率水平下各方法不同样本含量下的把握度（power）变化趋势。结果：在基于单组置信区间组合估计的几种MOVER方法中,MOVER Wilson计分法、MOVER Jeffreys法的I类错误率更接近事先设定的水平,尤其是在靠近0和100%两端时,MOVER Jeffreys法的I类错误率更优;除MOVER CP法,其他3种方法的把握度接近。结论：对于配对二项资料两组率差的置信区间估计,一般情况下（两组率在20%~80%范围内）,可选择MOVER Wilson计分法或MOVER Jeffreys法;当两组率靠近两端时,推荐使用MOVER Jeffreys法。

养多乐对酿酒葡萄“赤霞珠”果实品质的影响

以欧亚种酿酒葡萄"赤霞珠"为试材,研究了养多乐(Sugar mover)不同喷施次数与喷施浓度对酿酒葡萄果实品质的影响。结果表明:随着养多乐喷施次数的增加,其果实还原糖含量增加。喷施1 200倍液养多乐"赤霞珠"果实还原糖最高。随着养多乐喷施次数的增加,"赤霞珠"果实总花色素含量增加,连续喷施4次600倍养多乐效果最佳。喷施养多乐对"赤霞珠"果实总酚含量影响不大。

一种度量图像相似性和计算图编辑距离的新方法

由于在图编辑距离(GED)的计算中合理地为编辑操作定义代价函数相当困难,因此本文提出一种基于图结构的独立于代价函数定义的GED计算方法.它利用边缘方向直方图刻画图的结构,通过计算边缘方向直方图之间的距离来判断图的相异性,从而无需再定义代价函数.Earth Mover’s Distance(EMD)可以准确地计算直方图之间的距离,而且对于图在平面内的旋转所引起的直方图变化具有鲁棒性.为此,本文采用边缘方向直方图之间的EMD计算图编辑距离.将图像用图来表示,利用这种新的图编辑距离度量图像之间的相似性.实验结果表明本文提出的方法可以简单而有效地对图像进行聚类和分类,与基于谱序列计算图编辑距离的方法相比,可以更好地刻画图的结构差异.

基于小波分析与EMD的机车轴承故障诊断方法

机车滚动轴承发生故障时信号是非平稳的,其振动信号各频带的能量发生相应变化。针对故障发生在不同部位时,其振动信号能量分布重叠较少这一问题,将一种相似度度量earth mover’s distance(EMD)引入到故障部位分类中。该方法使用db3小波对轴承震动信号进行3层小波包分解,并计算第三层各结点能量作为该信号的特征向量。对得到的特征向量进行处理,计算特征向量间的EMD,根据EMD大小对特征向量间的相似度进行判断,并依此对故障部位进行定位。仿真结果表明,该方法诊断准确率达到98.75%,相较于传统KNN诊断方法在诊断准确率上提升了5%。该方法能够准确有效地诊断滚动轴承故障,可以应用到工业生产中。

基于Drupal构建国家击剑队视频管理平台的研究与实践

体育视频作为获取运动员比赛技战术特点的重要资料,已受到广泛重视。国家击剑队积累了大量各国运动员的比赛视频,但由于缺少科学的管理手段,造成了视频使用中的诸多困难。以内容管理平台Drupal为框架,搭建了国家击剑队视频管理平台。实际运用发现,该平台实现了视频资料的分类整理、网络共享,大大提高了工作效率。

Content Feature Extraction-based Hybrid Recommendation for Mobile Application Services

The number of mobile application services is showing an explosive growth trend,which makes it difficult for users to determine which ones are of interest.Especially,the new mobile application services are emerge continuously,most of them have not be rated when they need to be recommended to users.This is the typical problem of cold start in the field of collaborative filtering recommendation.This problem may makes it difficult for users to locate and acquire the services that they actually want,and the accuracy and novelty of service recommendations are also difficult to satisfy users.To solve this problem,a hybrid recommendation method for mobile application services based on content feature extraction is proposed in this paper.First,the proposed method in this paper extracts service content features through Natural Language Processing technologies such as word segmentation,part-of-speech tagging,and dependency parsing.It improves the accuracy of describing service attributes and the rationality of the method of calculating service similarity.Then,a language representation model called Bidirectional Encoder Representation from Transformers(BERT)is used to vectorize the content feature text,and an improved weighted word mover’s distance algorithm based on Term Frequency-Inverse Document Frequency(TFIDF-WMD)is used to calculate the similarity of mobile application services.Finally,the recommendation process is completed by combining the item-based collaborative filtering recommendation algorithm.The experimental results show that by using the proposed hybrid recommendation method presented in this paper,the cold start problem is alleviated to a certain extent,and the accuracy of the recommendation result has been significantly improved.

一种基于MOVER法中介效应置信区间的构造方法

目的提出一种基于MOVER法中介效应置信区间的构造方法,并与Sobel法、百分位数bootstrap法、偏差校正的百分位数bootstrap法、乘积分布法和Monte Carlo法进行比较。方法借鉴MOVER法的思想,提出一种新的构造中介效应置信区间的MMA方法。在不同参数设置下,通过计算机模拟,比较MMA法与现有方法的第一类错误率、检验效能、覆盖率及区间宽度。同时,将MMA法与现有方法应用于多发性硬化患者数据,对这些方法做进一步的比较。结果 MMA法在样本量大于25时均能控制第一类错误率,偏差校正的百分位数bootstrap法的第一类错误率可能会膨胀;除了偏差校正的百分位数bootstrap法,MMA法的检验效能明显高于其余方法;随着样本量的增大,MMA法的覆盖率逐渐趋近0.95;在部分模拟背景下,MMA法的区间宽度最窄。结论 MMA法是一种稳健且检验效能有明显提高的中介效应置信区间构造方法。

Industrial Policy and the Great Power Competition

Since the end of the Cold War,major powers have avoided direct military confrontation,wary of the devastating consequences of nuclear warfare.Yet the arms race and regional conflicts remain important forms of power competition.Nonetheless,in the era of globalization,rapid advancements in new technologies and industries have eclipsed the utility of geopolitical maneuvers and military competition.In today’s world,the great power competition goes beyond the arms race and GDP aggregates.What matters more is the race to innovate and apply new technologies through a complete range of industrial sectors.In a world where technology has a controlling influence,industrial security has become the linchpin of national security.As globalization reshapes the world’s industrial landscape,a country’s economic influence,military might and national security depend increasingly on its industrial structure.National security is contingent upon industrial strength.In the era of globalization,industrial policy,cutting-edge technologies and market size are dominant factors influencing a country’s competitive position.The essence of the great power competition are industrial policies that fully unlock a country’s industrial potential and implementation of the policies.

一种基于依存句法和WRD的句子相似计算方法

句子相似计算技术能够得到两个句子的相似关系,在敏感数据检测领域具有广阔的应用前景。因此,提出了一种融合依存句法和Word Rotator’s Distance语义距离的句子相似计算方法。通过依存句法分析句子结构并提取各种依存关系词集,计算对应依存关系词集间的Word Rotator’s Distance语义距离,并最终得到两个句子的相似指数。实验证明,该方法能够得到比较准确的句子相似计算结果,适用于多种敏感数据检测场景。

Behavior Targeting Based on Hierarchical Taxonomy Aggregation for Heterogeneous Online Shopping Applications

Behavior targeting(BT)based on individual web-browsing history has become more valuable in precision marketing for many companies through capturing users’interest and preference.It is common in practice that the behavior data collected from different online shopping applications are inconsistent since they are labelled by different item taxonomy,where the same behavior could have different representations and therefore analysis confusion arises.To address this issue,we propose a semantic similarity based strategy to transform the heterogeneous behavior extracted from deep packet inspection(DPI)data of a telecommunication operator into a unique standard one.The Word Mover’s Distance algorithm is exploited to evaluate the semantic similarity of the distributed representations of two web-browsing histories.Moreover,the architecture of the behavior targeting platform on Hadoop is implemented,which is capable of processing data with size of PB level every day.

Novel similarity measures for face representation based on local binary pattern

The successful face recognition based on local binary pattern(LBP)relies on the effective extraction of LBP features and the inferring of similarity between the extracted features.In this paper,we focus on the latter and propose two novel similarity measures for the local matching methods and the holistic matching methods respectively.One is Earth Mover's Distance with Hamming and Lp ground distance(EMD-HammingLp),which is a cross-bin dissimilarity measure for LBP histograms.The other is IMage Hamming Distance(IMHD),which is a dissimilarity measure for the whole LBP images.Experiments on FERET database show that the proposed two similarity measures outperform the state-of-the-art Chi-square similarity measure for extraction of LBP features.

Effect of Battery on Moving Properties of Cableless In-Piping Magnetic Actuator

This paper proposes a cableless in-piping magnetic actuator that exhibits a very high-speed locomotion into inner pipe of 8 mm. The cableless magnetic actuator is moved according to the vibration amplitude and resonance energy of a mass-spring system excited by using an electromagnetic force. The iron core size of the bobbin type electromagnet was roughly designed by computer simulation and then optimized experimentally. The proposed actuator incorporates an electrical inverter that directly transforms DC from button batteries into AC. The electrical DC-AC inverter incorporates a mass-spring system, a reed switch and a curved permanent magnet that switch under an electromagnetic force. The duty ratio is changed into this electrical inverter by changing the position of the curved magnet and the reed switch. Experimental result demonstrates that the cableless magnetic actuator was able to move horizontally at 471 m, and horizontal speed at 327 mm/s when Maxell SR621W silver-oxide button batteries were used.

对大功率LED驱动电路的研究

从原始到现代科技,人们的社会生活发生着翻天覆地的变化,人们对于光亮的追求从火的产生开始,火给人类带来了光明,但它会慢慢消耗,燃烧殆尽,后来在二十世纪八十年代初,美国的爱迪生发明了电灯,解决了光可以长期发光的问题,但白炽灯其本质还是利用物质的加热燃烧,即利用钨丝发光,钨丝的损耗会降低白炽灯的使用时间,所以人们又发明了LED灯,它是利用一种能够发光的半导体元器件发光,其相对稳定而耐用,慢慢的,LED设备被应用于社会发展的方方面面。针对大功率的LED发光设备驱动电路进行介绍和研究。