当前类别: 所有     查看:

搜 索 结 果

共有 310文章/程序   按 方式排序
论文/程序题目 作者 下载 主要内容介绍 Hits
Understanding the mesoscopic scaling patterns within cities
Lei Dong,Zhou Huang   Understanding quantitative relationships between urban elements is crucial for a wide range of applications. The observation at the macroscopic level demonstrates that the aggregated urban quantities (e.g., gross domestic product) scale systematically with population sizes across cities, also known as urban scaling laws. However, at the mesoscopic level, we lack an understanding of whether the simple scaling relationship holds within cities, which is a fundamental question regarding the spatial origin of scaling in urban systems. Here, by analyzing four extensive datasets covering millions of mobile phone users and urban facilities, we investigate the scaling phenomena within cities. We find that the mesoscopic infrastructure volume and socioeconomic activity scale sub- and super-linearly with the active population, respectively. For a same scaling phenomenon, however, the exponents vary in cities of similar population sizes. To explain these empirical observations, we propose a conceptual framework by considering the heterogeneous distributions of population and facilities, and the spatial interactions between them. Analytical and numerical results suggest that, despite the large number of complexities that influence urban activities, the simple interaction rules can effectively explain the observed regularity and heterogeneity in scaling behaviors within cities. 0
Coevolution of agent’s behavior and noise parameters in majority vote game on multi-layer networks
Jing Liu,Ying Fan   In many real-world systems, agents’behaviors are usually coupled with environmental changes. Tomodel their coevolutionary process, we study a non-equilibrium model known as majority vote modelcoupled with reaction-diffusion processes on a two-layer multiplex network. The dynamics of thenoise parameter in noise layer is related to the voting behavior and represents the increase and decreaseof social tension in the context of social dynamics. We perform Monte Carlo simulations andfinite-size scaling analysis in order to investigate the statistical behavior of the model. It is interesting tofindthat our coupling mechanism induces a continuous order–disorder phase transition on randomregular graphs, but the critical phenomenon disappears on square lattices. Besides, switching fromone sign of the spontaneous magnetization to the opposite is also observed near the critical region.Given specific parameters, the system may self-organize to a critical state from any initial conditions.In addition, a mean-field method is developed to study the properties of the phase transitionanalytically and the solutions are in good agreement with our numerical results qualitatively. 0
Unraveling Environmental Justice in Ambient PM2.5 Exposure in Beijing: A Big Data Approach
Yanyan Xu,Shan Jiang   Air pollution imposes significant environmental and health risks worldwide and is expected to deteriorate in the coming decade as cities expand. Measuring population exposure to air pollution is crucial to quantifying risks to public health. In this work, we introduce a big data analytics framework to model residents' stay and commuters' travel exposure to outdoor PM2.5 and evaluate their environmental justice, with Beijing as an example. Using mobile phone and census data, we first infer travel demand of the population to derive residents' stay activities in each analysis zone, and then focus on commuters and estimate their travel routes with a traffic assignment model. Based on air quality observations from monitoring stations and a spatial interpolation model, we estimate the outdoor PM2.5 concentrations at a 500-m grid level and map them to road networks. We then estimate the travel exposure for each road segment by multiplying the PM2.5 concentration and travel time spent on the road. By combining the estimated PM2.5 exposure and housing price harnessed from online housing transaction platforms, we discover that in the winter, Beijing commuters with low wealth level are exposed to 13% more PM2.5 per hour than those with high wealth level when staying at home, but exposed to less PM2.5 by 5% when commuting the same distance (due to lighter traffic congestion in suburban areas). We also find that the residents from the southern suburbs of Beijing have both lower level of wealth and higher stay- and travel- exposure to PM2.5, especially in the winter. These findings inform more equitable environmental mitigation policies for future sustainable development in Beijing. Finally, or the first time in the literature, we compare the results of exposure estimated from passive data with subjective measures of perceived air quality (PAQ) from a survey. The PAQ data was collected via a mobile-app. The comparison confirms consistencies in results and the advantages of the big data for air pollution exposure assessments. 0
PM2.5-GNN: A Domain Knowledge Enhanced Graph Neural Network For PM2.5 Forecasting
Shuo Wang,Yanran Li   在这个工作中,我们开发了一个PM2.5浓度预测的图神经网络模型。我们首先构建雾霾的传播网络,其中节点为城市,连边则根据两座城市之间的地理距离以及它们之间有没有高山来决定。除此之外,我们将天气信息融合进了连边特征之中。该模型能够融合领域知识给出相对高精度的预测。 0
Complex Network Classification with Convolutional Neural Network
Ruyue Xin,Jiang Zhang   Classifying large-scale networks into several categories and distinguishing them according to their fine structures is of great importance to several real-life applications. However, most studies on complex networks focus on the properties of a single network and seldom on classification, clustering, and comparison between different networks, in which the network is treated as a whole. Conventional methods can hardly be applied on networks directly due to the non-Euclidean properties of data. In this paper, we propose a novel framework of Complex Network Classifier (CNC) by integrating network embedding and convolutional neural network to tackle the problem of network classification. By training the classifier on synthetic complex network data, we show CNC can not only classify networks with high accuracy and robustness but can also extract the features of the networks automatically. We also compare our CNC with baseline methods on benchmark datasets, which shows that our method performs well on large-scale networks. 0
Using learning analytics to understand collective attention in language MOOCs
Shuang Zeng,Jingjing Zhang   Learning analytics (LA) has the potential to generate new insights into the complexities of learning behaviours in language massive open online courses (LMOOCs). In LA, the collective attention model takes an ecological system view of the dynamic process of unequal participation patterns in online and flexible learning environments. In this study, the‘Oral Communication for EFL Learners (spring)’on XuetangX was selected as an example with which to examine the allocation of learner attention in the context ofLMOOCs. The open-flow network of collective attention was used to model the dynamics of learning behaviours to understand how different cohorts of second language (L2)learners allocated their attention at the collective level. The results showed that what distinguished high-performing L2learners was related less to where they started withLMOOC resources or how much attention they allocated to certain learning units and more to the extent to which their attention could be maintained and circulated into other learning units. In addition, learners’s attention typically followed the pre-designed course structure each time they entered the online space. No learning resources offered in the selected LMOOC were found to dominate the collective attention flow, which suggested that L2 learners’ online engagement followed classroom learning patterns. The use of LA to understand the allocation of L2 attention at the collective level provides new perspectives on digital behaviour in LMOOCs, which may facilitate the design of cost-effective L2 resources that prevent learner overload in the information-rich age. 0
Investigating time, strength, and duration of measures in controlling the spread of COVID-19 using a networked meta-population model
Jiang Zhang,Lei Dong   2019年3月,新冠疫情自湖北武汉爆发,并迅速在全国传播。本论文利用大规模人类迁徙数据,构建了病毒在中国各大城市传播的meta-population模型。利用Neural ODE技术,我们首先根据实际疫情数据拟合了模型中包括R0的重要参数,然后推断了实际的感染人数。最后,模型探索了在不同干预政策的条件下,病毒的传播情况。 0
An interpretable deep-learning architecture of capsule networks for identifying cell-type gene expression programs from single-cell RNA-sequencing data
Lifei Wang,Rui Nie   随着单细胞RNA测序技术的发展和数据的积累,人们对单细胞类型标注深入到单细胞转录组分析的层次。本篇文章基于Capsule模型,提出了一种用于单细胞分类任务的具有可解释性的深度学习架构scCapsNet。通过对RNA数据的单一细胞类型分类任务的训练,scCapsNet可以捕获编码不同亚细胞类型的基因,并将RNA的表达特征融合进Capsule的参数矩阵中,并使得单一类型的识别成为可能,这一特征使得我们可以发现那些特定的基因调控单元,其中的基因在功能上密切相关,并彼此相互作用,但却具有不通的表达模式。 0
Human mobility in interest space and interactive random walk
Fanqi Zeng,Li Gong   Compared with the well-studied topic of human mobility in real geographic space, only a few studies focus on human mobility in virtual space, such as interests, knowledge, ideas, and so on. However, it relates to the issues like public opinion management, knowledge diffusion, and innovation. In this paper, we assume that the interests of a group of online users can span an Euclidean space which is called interest space, and the transfers of user interests can be modelled as Lévy Flight in the interest space. Considering the interaction between users, we assume that the random walkers are not independent but interacting with each other indirectly via the digital resources in the interest space. The proposed model in this paper successfully reproduced a set of scaling laws for describing the growth of attention flow networks of online communities, and obtaining similar ranges of users’ scaling exponents with empirical data. Further, we inferred parameters for describing the individual behaviours of the users according to the scaling laws of empirical attention flow network. Our model can not only provide theoretical understanding of human online behaviours but also has broad potential applications such as dissemination and public opinion management, online recommendation, etc. 0
A Topological Analysis of Trade Distance: Evidence from the Gravity Model and Complex Flow Networks
Zongning Wu,Hongbo Cai   As a classical trade model, the gravity model plays an important role in the trade policy-making process. However, the effect of physical distance fails to capture the effects of globalization and even ignores the multilateral resistance of trade. Here, we propose a general model describing the effective distance of trade according to multilateral trade paths information and the structure of the trade flow network. Quantifying effective trade distance aims to identify the hidden resistance information from trade networks data, and then describe trade barriers. The results show that flow distance, hybrid by multi-path constraint, and international trade network contribute to the forecasting of trade flows. Meanwhile, we also analyze the role of flow distance in international trade from two perspectives of network science and econometric model. At the econometric model level, flow distance can collapse to the predicting results of geographic distance in the proper time lagging variable, which can also reflect that flow distance contains geographical factors. At the international trade network level, community structure detection by flow distances and flow space embedding instructed that the formation of international trade networks is the tradeoff of international specialization in the trade value chain and geographical aggregation. The methodology and results can be generalized to the study of all kinds of product trade systems. 0
A General Deep Learning Framework for Network Reconstruction and Dynamics Learning
Zhang Zhang,Yi Zhao   本文章提出了一种通用的网络结构及其动力学重构技术。采用Gumbel softmax采样和图网络技术相结合,我们可以针对离散、连续、二值的时间序列数据进行网络和动力学重构,准确度超过了对比模型。 0
Modeling collective attention in online and flexible learning environments
Jingjing Zhang,Xiaodan Lou   Understanding how collective attention flow circulates amid an over-abundance of knowledge is a key to designing new and better forms of online and flexible learning experiences. This study adopted an open flow network model and the associated distance metrics to gain an understanding of collective attention flow using clickstream data in a massive open online course. Various patterns and dynamics of attention flow were identified and are discussed here in relation to learning performance. The results show that the effective accumulation, circulation, and dissipation of attention flow are important contributors to academic attainment. Understanding the patterns and dynamics of attention flow will allow us to design cost-effective learning resources to prevent learners from becoming overloaded. 0
Simple spatial scaling rules behind complex cities
Ruiqi Li,Lei Dong   城市自其出现以来,已然成为人类发展的重要驱动,目前全球有超过50%的人口居住在城市之中,超过80%的财富与90%的创新都产生于城市;但城市的发展同时也带来许多社会问题,例如污染、交通拥堵、各类犯罪等等。城市是典型的由多种元素构成的复杂系统,过去的研究往往只关注于城市的某一方面,而且目前对于城市的定量研究仍不够充分,难以定量预测城市中主要元素的空间分布(例如人口、道路、与社会经济相关的相互作用等等)。本文基于空间吸引和匹配生长机制提出了一个简单模型,首次揭示了城市中主要元素的空间标度律,而且各主要元素可以由统一的框架来解释,而这使得我们可以使用任何单一的分布来对其他的分布进行推断。此外文中提出的模型不但可以解释介观尺度的空间分布,还可以对跨城市的宏观超线性与亚线性标度律的起源作出一般性解释,并准确预测公里级的社会经济活动。而且我们的理论方法也突破了过去城市研究领域中全局平均场理论的假设,直接从增长演化的视角对城市进行建模。文章还提出一些全新的概念,例如活跃人口,这一概念可以解决过去对于人口分布形式究竟是指数还是幂律的部分争论,还可应用于城市街区安全评估。简言之,我们的工作提出了揭示城市元素间相互作用与演变的新视角和新方法,未来它将有广泛的应用场景。 0
The Hidden Flow Structure and Metric Space of Network Embedding Algorithms Based on Random Walks
Gu, Weiwei,Li Gong   Network embedding which encodes all vertices in a network as a set of numerical vectors in accordance with it’s local and global structures, has drawn widespread attention. Network embedding not only learns significant features of a network, such as the clustering and linking prediction but also learns the latent vector representation of the nodes which provides theoretical support for a variety of applications, such as visualization, link prediction, node classification, and recommendation. As the latest progress of the research, several algorithms based on random walks have been devised. Although those algorithms have drawn much attention for their high scores in learning efficiency and accuracy, there is still a lack of theoretical explanation, and the transparency of those algorithms has been doubted. Here, we propose an approach based on the open-flow network model to reveal the underlying flow structure and its hidden metric space of different random walk strategies on networks. We show that the essence of embedding based on random walks is the latent metric structure defined on the open-flow network. This not only deepens our understanding of random- walk-based embedding algorithms but also helps in finding new potential applications in network embedding. 0
The Atlas of Chinese World Wide Web Ecosystem Shaped by the Collective Attention Flows
Xiaodan Lou,You Li   我们采用CNNIC的调查数据,构造了由大量中国网民浏览行为形成的注意力流网络。根据流网络上的流距离度规,我们将整个注意力流网络嵌入到了一个高维空间中。在这个空间下,我们研究了整个中国互联网生态系统的分布情况。我们发现,网站会自发地形成4各区块。其中BAT三家大型网站构成了整个注意力的吞噬中心,它们占据了70%以上的注意力。另外,知识问答类、电子商务类、娱乐类、综合类网站分别聚集到地图中的不同位置上。 0
Population-weighted efficiency in transportation networks
Lei Dong,Ruiqi Li   Transportation efficiency is critical for the operation of cities and is attracting great attention worldwide. Improving the transportation efficiency can not only decrease energy consumption, reduce carbon emissions, but also accelerate people’s interactions, which will become more and more important for sustainable urban living. Generally, traffic conditions in less-developed countries are not so good due to the undeveloped economy and road networks, while this issue is rarely studied before, because traditional survey data in these areas are scarce. Nowadays, with the development of ubiquitous mobile phone data, we can explore the transportation efficiency in a new way. In this paper, based on users’ call detailed records (CDRs), we propose an indicator named population-weighted efficiency (PWE) to quantitatively measure the efficiency of the transportation networks. PWE can provide insights into transportation infrastructure development, according to which we identify dozens of inefficient routes at both the intra- and inter-city levels, which are verified by several ongoing construction projects in Senegal. In addition, we compare PWE with excess commuting indices, and the fitting result of PWE is better than excess commuting index, which also proves the validity of our method. 本篇文章运用人类移动数据定义了一种由人口流量调节的道路疏运效率指标,并用这个指标分析了塞内加尔的城市道路网络,挖掘出了一些低效率的路段。同时,文章还为每个城市计算了一个交通运输能力指数,对塞内加尔的不同城市做了系统性的比较。这套指标的好处在于它不依赖于官方提供的道路信息,只需要运用手机通讯数据以及Google提供的地图查询服务就可以完成计算。因而这套方法特别适用于那些经济落后地区,以及调查数据缺失的地区。 0
A Geometric Representation of Collective Attention Flows
Peiteng Shi,Xiaohan Huang   With the fast development of Internet and WWW, “information overload” has become an overwhelming problem, and collective attention of users will play a more important role nowadays. As a result, knowing how collective attention distributes and flows among different websites is the first step to understand the underlying dynamics of attention on WWW. In this paper, we propose a method to embed a large number of web sites into a high dimensional Euclidean space according to the novel concept of flow distance, which both considers connection topology between sites and collective click behaviors of users. With this geometric representation, we visualize the attention flow in the data set of Indiana university clickstream over one day. It turns out that all the websites can be embedded into a 20 dimensional ball, in which, close sites are always visited by users sequentially. The distributions of websites, attention flows, and dissipations can be divided into three spherical crowns (core, interim, and periphery). 20% popular sites (Google.com, Myspace.com, Facebook.com, etc.) attracting 75% attention flows with only 55% dissipations (log off users) locate in the central layer with the radius 4.1. While 60% sites attracting only about 22% traffics with almost 38% dissipations locate in the middle area with radius between 4.1 and 6.3. Other 20% sites are far from the central area. All the cumulative distributions of variables can be well fitted by “S”-shaped curves. And the patterns are stable across different periods. Thus, the overall distribution and the dynamics of collective attention on websites can be well exhibited by this geometric representation. 本篇文章通过点击流数据构建流网络,计算了任意两个网站的流距离,它能够反映两个网站相互联系的紧密程度。根据这些流距离,我们将所有的网站嵌入到一个20维空间中,于是我们可以看到不同的网站具有了不同的生态位。注意力流和网站的分布形成了三层的洋葱结构,最内层少数几个网站占据了绝大部分流量,中层有大多数网站,流量却很小,最外层少数网站有少数流量。 0
Scaling behaviours in the growth of networked systems and their geometric origins
Jiang Zhang,Xintong Li   In many networked systems (cities, online communities), links (or interactions) grow faster than nodes, as well, the diversity of nodes grow slower than nodes. We build a simple random network model based on geometric matching mechanism to reproduce both phenomena. The extensive model is further applied to model the distribution of natural cities. 在很多网路系统(在线社区、城市)中,连边(相互作用)总是会比节点以更快速度的增长。与此同时,节点的类别多样性会比节点以更慢的速度增长。本文章提出了一个简单的随机几何网络增长模型,同时给出了连边超线性生长和多样性亚线性生长的现象。通过改进该模型,我们还可以模拟城市系统的生长和分布。 0
Open Flow Distances on Open Flow Networks
Liangzhu Guo,Xiaodan Lou   Open flow network is a special weighted directed graph in which weighted links are flows, and the flows are in balance. We define a new set of distance metrics, which measure the average length of particles flow from i to j. Based on the distances, we discuss the calculation of trophic levels of specied on energetic food webs, the centrality of nodes, and the industrial clustering problem on input-ouput networks, etc. We also compare the new distances with old distances on graph. 开放流网络是一种特殊的加权有向网,其中加权连边表示流量,同时节点满足流平衡。在这种网络上,我们定义了一组流距离,即流子沿连边流动从i到j经历的平均路径长度。在此基础上,我们讨论了食物网的营养级计算、投入产出网上的节点中心度、产业聚类等问题,我们还比较了新的距离与其它网络距离。 0
Maximum Entropy for the International Division of Labor
Hongmei Lei,Ying Chen   As a result of the international division of labor, the trade value distribution on different products substantiated by international trade flows can be regarded as one country’s strategy for competition. Each country wants to diversify their investments on different products as well as make profits as possible as they can. We build a model based on maximum entropy principle to reproduce the distribution curves of countries. The results show that almost all countries' export share distributions can be explained by the maximum entropy model if the constraints are properly selected. 我们可以将每个国家在不同产品上的出口份额看作是一种面向国际市场的劳动分工策略。每个国家都在尽量多样化自己的出口多样性的同时牟取最大的经济利益。本文提出了一个最大化熵模型以解释各个国家在不同产品上的出口份额分布曲线。在合适地选择了最大化熵的约束条件后,我们成功地用一个单参数最大熵模型较好地拟合了100多个国家的出口分布曲线。 0
  
第1页/ 共16页
下一页
     全部 类别中查询