Role of modular and hierarchical structure in making networks dynamically stable

Size: px
Start display at page:

Download "Role of modular and hierarchical structure in making networks dynamically stable"

Transcription

1 Role of modular and hierarchical structure in making networks dynamically stable Raj Kumar Pan and Sitabhra Sinha The Institute of Mathematical Sciences, C.I.T. Campus, Taramani, Chennai India (Dated: August 7, 26) According to the May-Wigner theorem, increasing the complexity of random networks results in their dynamical instability. However, the prevalence of complex networks in nature, where they necessarily have to be robust to survive, appears to contradict this theoretical result. A possible solution to this apparent paradox maybe through the introduction of certain structural features of real-life networks, in particular, modularity and hierarchical levels. In this paper, we first show that the existence of these structures in an otherwise random network will make it more unstable. Next, we introduce the realistic constraint that every link has an associated cost, and, find that modular networks are indeed more stable than homogeneous networks. Increasing modularity in such networks results in the appearance of large number of hubs and heterogeneous degree distribution, a general property shared by many networks including scale-free networks. Our results provide a dynamical setting for explaining the ubiquity of such networks in reality. INTRODUCTION In recent times, the study of networks has received a lot of attention from physicists, biologists and social scientists [, 2]. The realization that many complex networks in nature and society share certain universal features, e.g., scale-free degree distribution, clustering, modularity, hierarchy, etc., has led to detailed investigation into the structure of such networks. However, much of the work to date has been focussed on the static aspects of networks, in particular, on the statistical properties of their connection topology. But most networks have associated dynamics, with the node properties evolving over time. A question of obvious significance is whether there is a relation between the stability of the network dynamics and its static aspects, e.g., the particular arrangement of the network connections. A network is said to be dynamically stable if small perturbations at a node quickly decay and are, therefore, unable to spread to the rest of the network. It has often been argued that complex networks are more stable if they have a larger number of nodes and links. Such assertions are partly based on field-observations by ecologists that have found more diverse and strongly connected ecosystems to be much more robust than their smaller, weakly connected counterparts [3]. However, theoretical work on the stability of model networks have tended to conclude the opposite. In particular, according to the May-Wigner theorem [4] for random networks, increasing the complexity (as measured by the number of nodes, density of connections and range of interaction strengths) always leads to decreased stability. One of the main objections against this result is that it is based on the study of networks whose connection topology shows none of the structures that are seen in real life networks, in particular, modularity and hierarchy. A network is said to be modular if it can be decomposed into sub-networks, such that there are significantly more connections within elements belonging to the same sub-network compared to that between elements belonging to different sub networks (Fig., top). Examples of modular networks include the neural network of the worm C. Elegans, consisting of 32 neurons and 27 synaptic connections between them (Fig. 2). On the other hand, many other networks, e.g., food webs or the communication structure among the employees of a typical company, show hierarchy [5]. A network has a hierarchical structure if the nodes are ordered into a certain set of layers with inter-layer connections occurring almost exclusively between adjacent layers, according to the pre-designated ordering (Fig., bottom). The twin features of hierarchy and modularity have been seen in a large number of empirical networks [6 9]. In this paper, we have undertaken a detailed study of the relation between the dynamical stability and network complexity for a simple network model where the modularity and hierarchy can be varied in a controlled manner. Initially, we observe the transition from stability to instability for randomly assembled networks, as the parameter controlling the range of interaction strengths is varied. By introducing various levels of hierarchy and modular structure, we conclude that both of these properties actually increase the instability of an otherwise random network. As this contradicts the observation that modularity and hierarchy are observed in many real-life networks, which necessarily have to be robust to survive ever-present environmental fluctuations, we impose further structure on the random networks by implementing the realistic constraint that every link in the network has a cost associated with it. In other words, the network assembly process tries to minimize the total number of links, L total. In this case, we find that the optimal network that minimizes L total while increasing its dynamical stability, will have a modular structure. We conclude with a short discussion of the implications of these results for the widespread occurrence of networks with multiple hubs and heterogeneous degree distribution in nature and society.

2 2 r 2 r 2 FIG. 3: (Left) Schematic diagram of the hierarchical modular network model, with the modules occurring at the various hierarchical levels indicated by ellipses, and (right) the corresponding adjacency matrix(right). FIG. : (Top left) Schematic diagram of a modular network, with modules demarcated by broken circles, with the corresponding adjacency matrix (top right) demonstrating the modularity in its approximately block diagonal structure. (Bottom left) Schematic diagram of a hierarchical network showing the characteristic layered structure, and the corresponding adjacency matrix (bottom right) nz = 27 FIG. 2: Synaptic connectivity matrix of the C. Elegans neural network. The lines are drawn as visual aids for identifying the modules. THE MODEL To look at the dynamical stability of a network, we consider the linear stability of an arbitrarily chosen fixed point of the network dynamics. If the network has N nodes, each node i being associated with a variable x i, then the time-evolution of the system is characterized by a N dimensional dynamical system ẋ = f(x), where f is a general nonlinear function. We investigate the stability around the fixed point x, i.e., if x = x +δx, then we study how the small perturbation δx grows or decays with time according to the equation δx = Jx, where J is the Jacobian matrix representing the interactions among the nodes: J ij = f i / x j x. As we are interested in the instability induced in the network, rather than the intrinsic instability of individual unconnected nodes, we can (without much loss of generality) set the diagonal element J ii =. This implies that, in the absence of any connections among the nodes, they are selfregulating, i.e., the fixed point x is stable. The behavior of the perturbation is determined by the largest real part, Re(λ) max, of the eigenvalues of J. If Re(λ) max >, an initially small perturbation will grow exponentially with time, and the system will be rapidly dislodged from the equilibrium state x. The relation between the dynamical properties and the static structure of the network is provided by its adjacency matrix A ij, which is if there is a connection between nodes i and j, and otherwise. There is a direct correspondence between the nature of the matrices J (specifying the dynamical behavior of perturbation) and A (which determines the structure of the underlying directed network), because A ij = implies J ij =. In our model, we have generated J ij from A ij by randomly choosing the non-zero elements from a Gaussian distribution with zero mean and variance σ 2. Our study is, therefore, conducted on a random statistical ensemble of J matrices for networks with an underlying structure decided by the adjacency matrix A. If J is an unstructured random matrix, then the largest real part of its eigenvalues, Re(λ) max NCσ 2, where C is the connectivity, i.e., the probability that there is a link between any pair of nodes in the network, and σ (defined above) is taken to be a measure for the range of interaction strengths [4]. When any of the parameters, N, C, or σ, is increased, there is a transition from stability to instability. This result, implying that complexity promotes instability, has been shown to be remarkably robust with respect to various generalizations [ 3]. Note that, the random matrices investigated in many previous studies correspond to Erdös-Rényi (ER) random graphs [4]. As mentioned before, real world networks differ from these purely random graphs [, 2] in having a particular structure among the arrangement of connections between nodes.

3 3 We introduce modular and hierarchical structure in an otherwise random network by dividing it into m modules (Fig. 3). This is done by generating the adjacency matrix A such that, the connectivity within each module, C intra =, is (in general) different from the connectivity between two modules belonging to the same hierarchical level, C () inter = r, where r is the parameter controlling the degree of modularity in the network. Note that, r = corresponds to the previously considered case of a ER random network, while r = corresponds to m disconnected random networks. Therefore, by changing r we can switch between completely modular (r = ) and completely homogeneous (r = ) networks. If two modules are separated by l levels in the hierarchy, then the connectivity between these two modules is C (l) inter = rl. The stochastic construction process of our network model, along with the ability to vary modularity (r) independent of hierarchy (l), makes it much more general than the deterministic model of hierarchical modular networks proposed by Ravasz & Barabasi [5]. In addition, unlike in this previous study, P(k), the distribution of the vertex degree k (the number of links associated with a node), need not be scale-free. In fact, it turns out that the criterion for hierarchical modularity given in Ref. [5], namely, that the clustering for vertices of degree k, C(k) /k, is crucially dependent on the scale-free degree distribution of the entire network. Thus, for a network having a different nature of P(k), e.g., the C. Elegans network which has an exponential degree distribution, the above criterion will indicate absence of hierarchy or modularity even if such structures do exist. Also, in our model, a module is defined to necessarily have more than one node, in contrast to a recent study where even a single node is considered to be a module [6]. This is a significant difference, as the latter criterion can lead to the identification of a set of disconnected nodes as a modular network, when in fact there is no network at all. RESULTS We look at the effect of modularity (r) and hierarchy (l) on the stability of the model network, by observing how changing their values affect the transition from stability to instability as one of the network parameters N,C,σ 2 is varied. As mentioned before, for ER networks it is known that if N,C are fixed, then the critical value of σ at which the transition to instability occurs is σ c / NC. Therefore, we check whether σ c changes with r and l. We arrange the network into hierarchical levels, such that at each level every module can be split further into a number of smaller modules; in the present study, a module has been divided into two modules at each level. Therefore, increasing the hierarchical level l results in increasing the number of modules; this is to be P Stability Variance, σ σ 2 FIG. 4: The probability of stability of a random hierarchical modular network (with N = 28 nodes and overall connectivity C =.2) varying with the measure of average interaction strength, σ, observed as a function of increasing hierarchy (left) and modularity (right). The weights corresponding to the links are chosen from a normal(, σ 2 ) distribution. At each hierarchical level, every module is split further into two modules. (Left) The N nodes are divided equally among 2 l modules (corresponding to l hierarchical levels) with r =.. Starting from top, the four different curves correspond to l =,, 2, 3. Increasing l shifts the transition to a smaller value of σ 2, implying that increasing hierarchy is destabilizing. (Right) The N nodes are divided among modules arranged in two hierarchical layers (l = 2), such that nodes belonging to the same module in a given hierarchical layer have r times more connections amongst them, than with nodes belonging to a different module. Starting from top, the different curves correspond to r =,.5,.,. Increasing r from r = (isolated modules) to r = (completely homogeneous network) shifts the transition to a larger value of σ 2, indicating that increasing modularity is destabilizing. contrasted with the operation of increasing the number of modules without any associated hierarchy that we shall consider later. As seen from Fig. 4, increasing either the hierarchy or modularity results in decreased stability of the network. This result appears contradictory to the observation that many real-life networks exhibit modularity to a greater or lesser extent. To understand this, we focus on the role of modularity in the dynamical stability of networks. For this purpose, we consider a simpler model, one where the random network having N nodes is divided into m modules, without any hierarchical structure. We divide a network of size N equally into m modules where the modules are connected to each other by a single link. This is effectively the same as joining m random networks of size N/m with single links to form a bigger network of size N. One can now compare the stability of this network against the homogeneous ER network with N nodes (subject to the constraint that both systems have the same average degree, < k >= NC), and thus investigate the effect of modularity on stability. Consistent with the results obtained for hierarchical modular networks, we observe that increasing m decreases stability (Fig. 5). However, from the May-Wigner theorem, we should have naively expected the stability-instability transition to be unchanged, since for any m, σ c = / NC is the same. P stability r = r =. r =.5 r =

4 N = 256 P{ Re ( λ ) max } P stability σ 2 m = m = 2 m = 4 m = 8 P { Re ( λ ) max } N = 28 N = 64 N = Re ( λ ) max Re ( λ ) max FIG. 5: Probability distribution of the largest real-part of eigenvalue for random network shown as a function of modularity, m. The network consists of N = 256 nodes divided equally into m modules where only a single undirected link exists between two modules. The connectivity of the entire network is C =.2 and the weights of the links are chosen from a normal(, σ 2 ) distribution with σ 2 =.3. The inset shows the probability of stability for random networks as a function of modularity. Increasing the modularity causes the transition at a lower value of σ 2, thus indicating increasing modularity decreases stability for random networks. To solve this apparent puzzle we first consider how the stability of a random network is affected by its size, N, when the average degree (< k >= NC) is kept constant. As seen in Fig. 6, decreasing N causes the distribution of the largest real part of the eigenvalues Re(λ) max to become more long-tailed. As a result, close to the stabilityinstability transition region, the probability of Re(λ) max having a positive value is slightly higher for networks with smaller N. Thus, when we consider the statistics of extremes, it is clear that close to the critical value of σ c, the smaller networks are more likely to have instability inducing fluctuations compared to the larger networks. Now we consider the fact that when a network of size N is split into m modules, the probability of stability of the entire network is decided by the probability of stability of the most unstable of the m modules of size N/m, ignoring the small additional effect of inter-modular connections. In other words, even if only one of the modules is unstable, the entire network will be considered unstable. Thus, the stability of the network of size N is decided by randomly drawing m values from the distribution of Re(λ) max for network of size N/m. As pointed out before, the longer tail of the distribution for smaller networks implies that it is now much more likely to obtain a positive value of Re(λ) max (especially in view of the m multiple drawings), than for the case of a homogeneous network of size N. This explains why the modular network is more unstable than the homogeneous network, FIG. 6: Probability distribution of the largest real-part of eigenvalue for ER random networks shown as a function of size N. The average degree < k >= 3, and the weights of the links are chosen from a normal(, σ 2 ) distribution with σ 2 =.3. The product NCσ 2 is same for all the networks. even though the two systems have the same total number of links. Although the results shown here are for modules connected to each other through a single link, the conclusion holds for the more general case of multiple inter-modular links (as seen in the hierarchical modular network results for r presented above). However, this does not answer our original question of why we see modular networks in nature at all. This seems all the more difficult to answer when we realize that modularity also increases the average path length of a network, thereby increasing communication time within the network. Therefore, modular networks can be advantageous over other connection topologies only if, on introducing a certain important constraint usually imposed in real-life networks, modularity can actually turn out to be stabilizing. A candidate for such a constraint is the link cost [7], i.e., the cost involved in introducing each additional link in a network, e.g., in the case of airline networks. Although the fastest network would be one where direct point-to-point flights connect every pair of airports, such a system would be prohibitively costly as each flight route would be expensive to maintain. In reality, therefore, one observes the existence of airline hubs, which acts as transit points for passengers arriving from and going to other airports. To see this constraint in terms of our model, we first note that a random network is distance minimizing but not cost minimizing, when cost is measured as being proportional to the total number of links. This is because a minimum number of links ( NlnN) is required to ensure that a random network is connected, i.e., there are no isolated sub-graphs [8]. It is easy to see that introducing the constraint of least cost (i.e., minimizing L total ) while requiring shortest communication in terms of average path length, leads

5 5.5 stability.8 P{ Re ( λ ) max } P σ.2 m= m=2. m=4 m=8 FIG. 7: Networks with star topology: star network (left) and clustered star network (right). i=2 where the hub node is labelled (Jii = ). If Jj = ±, PN then i=2 Ji Ji is just the displacement due to a dimensional random walk of length N. Splitting the star network into m modules, such that each module is connected to two others through links between the corresponding hub nodes, we have for each module, p Re(λ)max N/m. For the entire p system of N nodes also, the largest eigenvalue is N/m, because the effect of the few additional (i.e., inter-modular) links to connect the modules to each other is negligible. Therefore, we can increase the stability by decreasing the size of a star network through dividing the entire network into a connected set of small star-shaped modules. (Fig. 8). However, as a star network is a very special case, to generalize it we put additional links between the non-hub nodes. We call this generalized version of a star network with a few additional connections as a clustered star network (Fig. 7, right). Even for the clustered star networks, we find that stability increases with increasing number of Re ( λ ) max FIG. 8: Probability distribution of the largest real-part of eigenvalues for star network shown as a function of modularity, m. The network consists of N = 256 nodes divided equally into m modules with a single bidirectional link between two modules. The weights of the links are chosen from a normal(, σ 2 ) distribution with σ 2 =.. The adjoining figure of a network shows how the star-shaped modules are connected to each other. The inset shows the probability of stability for clustered star networks as a function of modularity. Increasing the modularity causes the transition to occur at a higher value of σ 2, indicating that modularity increases stability for clustered star networks. Im (λ) to a star-shaped connection topology (Fig. 7, left). We define a star network as one where there is a single node acting as the hub, to which all the other nodes are connected, there being no connections between the non-hub nodes. In a star network, the total number of connections required to make it connected is 2(N ). The average degree (links per node), < k > 2 for a connected star network is much less than that of a connected random network (< k > ln N ). We now observe that for star networks, modularity is stabilizing. The largest real part of the eigenvalues of a star network, Re(λ)max N, where N is the total number of nodes, in contrast to random networks, where Re(λ)max k. In other words, while the instability of random networks increases only with average degree, that of star networks increases with the size of the network itself. To see this, construct the corresponding Jacobian matrix J according to the prescription outlined before, and note that the eigenvalues of the matrix are v un ux λ = ± t Ji Ji,,,...,, 2 Re (λ) 2 Re (λ) FIG. 9: Eigenvalue plane for clustered star network for (left) m =, i.e., no modular structure and (right) m = 8, i.e., divided into 8 modules, where the modules are connected by undirected links between the hubs. A broken circle of radius centred at (-,) is shown for visual guidance. modules (Fig. 8, inset). The difference between the clustered star network and the random network is only in the presence of the hub nodes, as the other nodes are connected to each other randomly. The connectivity of the entire network is ensured by linking all the hub nodes of the modules with each other. The effect of modularity on the stability-instability transition is seen more clearly when we look at the eigenvalue plain of the clustered star network in both the modular and the non-modular case (Fig. 9). As is evident, increasing the modularity of the network causes the radius of the circle bounding all the eigenvalues to shrink. As a result, the entire network becomes more stable when it is divided into a large number of modules.

6 6 CONCLUSION In this paper we have shown that, although hierarchy and modularity in random networks are destabilizing, when we introduce real-life constraints, such as a cost per link, modular networks turn out to be stabilizing. Note that, an alternative way of stating this is that stable networks will exhibit multiple hubs and will have a heterogeneous degree distribution. Many types of networks, including scale-free networks [9], can be seen as special cases of this general criterion. Therefore, by introducing dynamical considerations along with a link-cost constraint, we can understand the large-scale occurrence of such networks in the natural and social world. [] M. E. J. Newman, SIAM Review 45, 67 (23). [2] R. Albert and A.-L. Barabási, Rev. Mod. Phys. 74, 47 (22). [3] C. S. Elton, The Ecology of Invasions by Animals and Plants, Methuen, London (958). [4] R. M. May, Stability and Complexity in Model Ecosystems, Princeton University Press, Princeton, NJ (973). [5] T. Hogg, B. A. Huberman and J. M. McGlade, Proc. Roy. Soc. Lond. B 237, 43 (989). [6] E. Ravasz, A. L. Somera, D. A. Mongru, Z. N. Oltvai and A.-L. Barabási, Science 297, 55 (22). [7] R. V. Solé and A. Munteanu, Europhys. Lett. 68, 7 (24). [8] P. Holme, M. Huss and H. Jeong, Bioinformatics 9, 532 (23). [9] A. E. Krause, K. A. Frank, D. M. Mason, R. U. Ulanowicz and W. W. Taylor, Nature 426, 282 (23). [] V. K. Jirsa and M. Ding, Phys. Rev. Lett. 93, 762 (24). [] S. Sinha and S. Sinha, Phys. Rev. E 7, 292(R) (25). [2] S. Sinha, Physica A 346, 47 (25). [3] M. Brede and S. Sinha, arxiv cond-mat/577 (25). [4] P. Erdös and A. Rényi, Publ. Math. Inst. Hung. Acad. Sci., Ser. A 5, 7 (96). [5] E. Ravasz and A.-L. Barabási, Phys. Rev. E 67, 262 (23). [6] E. A. Vaiano, J. M. McCoy and H. Lipson, Phys. Rev. Lett. 92, 887 (24). [7] N. Mathias and V. Gopal, Phys. Rev. E. 63, 27 (2). [8] B. Bollobas, Random Graphs, Cambridge University Press, Cambridge (2). [9] A.-L. Barabási and R. Albert, Science 286, 59 (999).

Higher order clustering coecients in Barabasi Albert networks

Higher order clustering coecients in Barabasi Albert networks Physica A 316 (2002) 688 694 www.elsevier.com/locate/physa Higher order clustering coecients in Barabasi Albert networks Agata Fronczak, Janusz A. Ho lyst, Maciej Jedynak, Julian Sienkiewicz Faculty of

More information

Smallest small-world network

Smallest small-world network Smallest small-world network Takashi Nishikawa, 1, * Adilson E. Motter, 1, Ying-Cheng Lai, 1,2 and Frank C. Hoppensteadt 1,2 1 Department of Mathematics, Center for Systems Science and Engineering Research,

More information

arxiv:cond-mat/ v1 [cond-mat.dis-nn] 3 Aug 2000

arxiv:cond-mat/ v1 [cond-mat.dis-nn] 3 Aug 2000 Error and attack tolerance of complex networks arxiv:cond-mat/0008064v1 [cond-mat.dis-nn] 3 Aug 2000 Réka Albert, Hawoong Jeong, Albert-László Barabási Department of Physics, University of Notre Dame,

More information

Small World Properties Generated by a New Algorithm Under Same Degree of All Nodes

Small World Properties Generated by a New Algorithm Under Same Degree of All Nodes Commun. Theor. Phys. (Beijing, China) 45 (2006) pp. 950 954 c International Academic Publishers Vol. 45, No. 5, May 15, 2006 Small World Properties Generated by a New Algorithm Under Same Degree of All

More information

Critical Phenomena in Complex Networks

Critical Phenomena in Complex Networks Critical Phenomena in Complex Networks Term essay for Physics 563: Phase Transitions and the Renormalization Group University of Illinois at Urbana-Champaign Vikyath Deviprasad Rao 11 May 2012 Abstract

More information

Lesson 4. Random graphs. Sergio Barbarossa. UPC - Barcelona - July 2008

Lesson 4. Random graphs. Sergio Barbarossa. UPC - Barcelona - July 2008 Lesson 4 Random graphs Sergio Barbarossa Graph models 1. Uncorrelated random graph (Erdős, Rényi) N nodes are connected through n edges which are chosen randomly from the possible configurations 2. Binomial

More information

Properties of Biological Networks

Properties of Biological Networks Properties of Biological Networks presented by: Ola Hamud June 12, 2013 Supervisor: Prof. Ron Pinter Based on: NETWORK BIOLOGY: UNDERSTANDING THE CELL S FUNCTIONAL ORGANIZATION By Albert-László Barabási

More information

Failure in Complex Social Networks

Failure in Complex Social Networks Journal of Mathematical Sociology, 33:64 68, 2009 Copyright # Taylor & Francis Group, LLC ISSN: 0022-250X print/1545-5874 online DOI: 10.1080/00222500802536988 Failure in Complex Social Networks Damon

More information

Constructing a G(N, p) Network

Constructing a G(N, p) Network Random Graph Theory Dr. Natarajan Meghanathan Professor Department of Computer Science Jackson State University, Jackson, MS E-mail: natarajan.meghanathan@jsums.edu Introduction At first inspection, most

More information

CSCI5070 Advanced Topics in Social Computing

CSCI5070 Advanced Topics in Social Computing CSCI5070 Advanced Topics in Social Computing Irwin King The Chinese University of Hong Kong king@cse.cuhk.edu.hk!! 2012 All Rights Reserved. Outline Graphs Origins Definition Spectral Properties Type of

More information

Dynamic network generative model

Dynamic network generative model Dynamic network generative model Habiba, Chayant Tantipathanananandh, Tanya Berger-Wolf University of Illinois at Chicago. In this work we present a statistical model for generating realistic dynamic networks

More information

Modelling weighted networks using connection count

Modelling weighted networks using connection count Home Search Collections Journals About Contact us My IOPscience Modelling weighted networks using connection count This article has been downloaded from IOPscience. Please scroll down to see the full text

More information

Constructing a G(N, p) Network

Constructing a G(N, p) Network Random Graph Theory Dr. Natarajan Meghanathan Associate Professor Department of Computer Science Jackson State University, Jackson, MS E-mail: natarajan.meghanathan@jsums.edu Introduction At first inspection,

More information

Search in weighted complex networks

Search in weighted complex networks Search in weighted complex networks Hari P. Thadakamalla, R. Albert 2 and S. R. T. Kumara Department of Industrial Engineering, The Pennsylvania State University, University Park, PA 6802 2 Department

More information

Complex Networks: Ubiquity, Importance and Implications. Alessandro Vespignani

Complex Networks: Ubiquity, Importance and Implications. Alessandro Vespignani Contribution : 2005 NAE Frontiers of Engineering Complex Networks: Ubiquity, Importance and Implications Alessandro Vespignani School of Informatics and Department of Physics, Indiana University, USA 1

More information

Response Network Emerging from Simple Perturbation

Response Network Emerging from Simple Perturbation Journal of the Korean Physical Society, Vol 44, No 3, March 2004, pp 628 632 Response Network Emerging from Simple Perturbation S-W Son, D-H Kim, Y-Y Ahn and H Jeong Department of Physics, Korea Advanced

More information

Complex Networks. Structure and Dynamics

Complex Networks. Structure and Dynamics Complex Networks Structure and Dynamics Ying-Cheng Lai Department of Mathematics and Statistics Department of Electrical Engineering Arizona State University Collaborators! Adilson E. Motter, now at Max-Planck

More information

Characteristics of Preferentially Attached Network Grown from. Small World

Characteristics of Preferentially Attached Network Grown from. Small World Characteristics of Preferentially Attached Network Grown from Small World Seungyoung Lee Graduate School of Innovation and Technology Management, Korea Advanced Institute of Science and Technology, Daejeon

More information

Graph Theory. Graph Theory. COURSE: Introduction to Biological Networks. Euler s Solution LECTURE 1: INTRODUCTION TO NETWORKS.

Graph Theory. Graph Theory. COURSE: Introduction to Biological Networks. Euler s Solution LECTURE 1: INTRODUCTION TO NETWORKS. Graph Theory COURSE: Introduction to Biological Networks LECTURE 1: INTRODUCTION TO NETWORKS Arun Krishnan Koenigsberg, Russia Is it possible to walk with a route that crosses each bridge exactly once,

More information

Universal Behavior of Load Distribution in Scale-free Networks

Universal Behavior of Load Distribution in Scale-free Networks Universal Behavior of Load Distribution in Scale-free Networks K.-I. Goh, B. Kahng, and D. Kim School of Physics and Center for Theoretical Physics, Seoul National University, Seoul 151-747, Korea (February

More information

CS-E5740. Complex Networks. Scale-free networks

CS-E5740. Complex Networks. Scale-free networks CS-E5740 Complex Networks Scale-free networks Course outline 1. Introduction (motivation, definitions, etc. ) 2. Static network models: random and small-world networks 3. Growing network models: scale-free

More information

Structural Analysis of Paper Citation and Co-Authorship Networks using Network Analysis Techniques

Structural Analysis of Paper Citation and Co-Authorship Networks using Network Analysis Techniques Structural Analysis of Paper Citation and Co-Authorship Networks using Network Analysis Techniques Kouhei Sugiyama, Hiroyuki Ohsaki and Makoto Imase Graduate School of Information Science and Technology,

More information

Network Theory: Social, Mythological and Fictional Networks. Math 485, Spring 2018 (Midterm Report) Christina West, Taylor Martins, Yihe Hao

Network Theory: Social, Mythological and Fictional Networks. Math 485, Spring 2018 (Midterm Report) Christina West, Taylor Martins, Yihe Hao Network Theory: Social, Mythological and Fictional Networks Math 485, Spring 2018 (Midterm Report) Christina West, Taylor Martins, Yihe Hao Abstract: Comparative mythology is a largely qualitative and

More information

The Establishment Game. Motivation

The Establishment Game. Motivation Motivation Motivation The network models so far neglect the attributes, traits of the nodes. A node can represent anything, people, web pages, computers, etc. Motivation The network models so far neglect

More information

arxiv:cond-mat/ v1 21 Oct 1999

arxiv:cond-mat/ v1 21 Oct 1999 Emergence of Scaling in Random Networks Albert-László Barabási and Réka Albert Department of Physics, University of Notre-Dame, Notre-Dame, IN 46556 arxiv:cond-mat/9910332 v1 21 Oct 1999 Systems as diverse

More information

Statistical Analysis of the Metropolitan Seoul Subway System: Network Structure and Passenger Flows arxiv: v1 [physics.soc-ph] 12 May 2008

Statistical Analysis of the Metropolitan Seoul Subway System: Network Structure and Passenger Flows arxiv: v1 [physics.soc-ph] 12 May 2008 Statistical Analysis of the Metropolitan Seoul Subway System: Network Structure and Passenger Flows arxiv:0805.1712v1 [physics.soc-ph] 12 May 2008 Keumsook Lee a,b Woo-Sung Jung c Jong Soo Park d M. Y.

More information

Introduction to network metrics

Introduction to network metrics Universitat Politècnica de Catalunya Version 0.5 Complex and Social Networks (2018-2019) Master in Innovation and Research in Informatics (MIRI) Instructors Argimiro Arratia, argimiro@cs.upc.edu, http://www.cs.upc.edu/~argimiro/

More information

Onset of traffic congestion in complex networks

Onset of traffic congestion in complex networks Onset of traffic congestion in complex networks Liang Zhao, 1,2 Ying-Cheng Lai, 1,3 Kwangho Park, 1 and Nong Ye 4 1 Department of Mathematics and Statistics, Arizona State University, Tempe, Arizona 85287,

More information

The coupling effect on VRTP of SIR epidemics in Scale- Free Networks

The coupling effect on VRTP of SIR epidemics in Scale- Free Networks The coupling effect on VRTP of SIR epidemics in Scale- Free Networks Kiseong Kim iames@kaist.ac.kr Sangyeon Lee lsy5518@kaist.ac.kr Kwang-Hyung Lee* khlee@kaist.ac.kr Doheon Lee* dhlee@kaist.ac.kr ABSTRACT

More information

M.E.J. Newman: Models of the Small World

M.E.J. Newman: Models of the Small World A Review Adaptive Informatics Research Centre Helsinki University of Technology November 7, 2007 Vocabulary N number of nodes of the graph l average distance between nodes D diameter of the graph d is

More information

(Social) Networks Analysis III. Prof. Dr. Daning Hu Department of Informatics University of Zurich

(Social) Networks Analysis III. Prof. Dr. Daning Hu Department of Informatics University of Zurich (Social) Networks Analysis III Prof. Dr. Daning Hu Department of Informatics University of Zurich Outline Network Topological Analysis Network Models Random Networks Small-World Networks Scale-Free Networks

More information

An Evolving Network Model With Local-World Structure

An Evolving Network Model With Local-World Structure The Eighth International Symposium on Operations Research and Its Applications (ISORA 09) Zhangjiajie, China, September 20 22, 2009 Copyright 2009 ORSC & APORC, pp. 47 423 An Evolving Network odel With

More information

Global dynamic routing for scale-free networks

Global dynamic routing for scale-free networks Global dynamic routing for scale-free networks Xiang Ling, Mao-Bin Hu,* Rui Jiang, and Qing-Song Wu School of Engineering Science, University of Science and Technology of China, Hefei 230026, People s

More information

arxiv:cond-mat/ v2 [cond-mat.stat-mech] 1 Sep 2002

arxiv:cond-mat/ v2 [cond-mat.stat-mech] 1 Sep 2002 APS/123-QED Hierarchical organization in complex networs arxiv:cond-mat/0206130v2 [cond-mat.stat-mech] 1 Sep 2002 Erzsébet Ravasz and Albert-László Barabási Department of Physics, 225 Nieuwland Science

More information

Centrality Book. cohesion.

Centrality Book. cohesion. Cohesion The graph-theoretic terms discussed in the previous chapter have very specific and concrete meanings which are highly shared across the field of graph theory and other fields like social network

More information

Example for calculation of clustering coefficient Node N 1 has 8 neighbors (red arrows) There are 12 connectivities among neighbors (blue arrows)

Example for calculation of clustering coefficient Node N 1 has 8 neighbors (red arrows) There are 12 connectivities among neighbors (blue arrows) Example for calculation of clustering coefficient Node N 1 has 8 neighbors (red arrows) There are 12 connectivities among neighbors (blue arrows) Average clustering coefficient of a graph Overall measure

More information

BUBBLE RAP: Social-Based Forwarding in Delay-Tolerant Networks

BUBBLE RAP: Social-Based Forwarding in Delay-Tolerant Networks 1 BUBBLE RAP: Social-Based Forwarding in Delay-Tolerant Networks Pan Hui, Jon Crowcroft, Eiko Yoneki Presented By: Shaymaa Khater 2 Outline Introduction. Goals. Data Sets. Community Detection Algorithms

More information

Deterministic small-world communication networks

Deterministic small-world communication networks Information Processing Letters 76 (2000) 83 90 Deterministic small-world communication networks Francesc Comellas a,,javierozón a, Joseph G. Peters b a Departament de Matemàtica Aplicada i Telemàtica,

More information

Attack Vulnerability of Network with Duplication-Divergence Mechanism

Attack Vulnerability of Network with Duplication-Divergence Mechanism Commun. Theor. Phys. (Beijing, China) 48 (2007) pp. 754 758 c International Academic Publishers Vol. 48, No. 4, October 5, 2007 Attack Vulnerability of Network with Duplication-Divergence Mechanism WANG

More information

The Complex Network Phenomena. and Their Origin

The Complex Network Phenomena. and Their Origin The Complex Network Phenomena and Their Origin An Annotated Bibliography ESL 33C 003180159 Instructor: Gerriet Janssen Match 18, 2004 Introduction A coupled system can be described as a complex network,

More information

TELCOM2125: Network Science and Analysis

TELCOM2125: Network Science and Analysis School of Information Sciences University of Pittsburgh TELCOM2125: Network Science and Analysis Konstantinos Pelechrinis Spring 2015 Figures are taken from: M.E.J. Newman, Networks: An Introduction 2

More information

Erdős-Rényi Model for network formation

Erdős-Rényi Model for network formation Network Science: Erdős-Rényi Model for network formation Ozalp Babaoglu Dipartimento di Informatica Scienza e Ingegneria Università di Bologna www.cs.unibo.it/babaoglu/ Why model? Simpler representation

More information

Course Introduction / Review of Fundamentals of Graph Theory

Course Introduction / Review of Fundamentals of Graph Theory Course Introduction / Review of Fundamentals of Graph Theory Hiroki Sayama sayama@binghamton.edu Rise of Network Science (From Barabasi 2010) 2 Network models Many discrete parts involved Classic mean-field

More information

Attacks and Cascades in Complex Networks

Attacks and Cascades in Complex Networks Attacks and Cascades in Complex Networks Ying-Cheng Lai 1, Adilson E. Motter 2, and Takashi Nishikawa 3 1 Department of Mathematics and Statistics, Department of Electrical Engineering, Arizona State University,

More information

Research on Community Structure in Bus Transport Networks

Research on Community Structure in Bus Transport Networks Commun. Theor. Phys. (Beijing, China) 52 (2009) pp. 1025 1030 c Chinese Physical Society and IOP Publishing Ltd Vol. 52, No. 6, December 15, 2009 Research on Community Structure in Bus Transport Networks

More information

Basics of Network Analysis

Basics of Network Analysis Basics of Network Analysis Hiroki Sayama sayama@binghamton.edu Graph = Network G(V, E): graph (network) V: vertices (nodes), E: edges (links) 1 Nodes = 1, 2, 3, 4, 5 2 3 Links = 12, 13, 15, 23,

More information

MAE 298, Lecture 9 April 30, Web search and decentralized search on small-worlds

MAE 298, Lecture 9 April 30, Web search and decentralized search on small-worlds MAE 298, Lecture 9 April 30, 2007 Web search and decentralized search on small-worlds Search for information Assume some resource of interest is stored at the vertices of a network: Web pages Files in

More information

Cascading failures in complex networks with community structure

Cascading failures in complex networks with community structure International Journal of Modern Physics C Vol. 25, No. 5 (2014) 1440005 (10 pages) #.c World Scienti c Publishing Company DOI: 10.1142/S0129183114400051 Cascading failures in complex networks with community

More information

CSCI5070 Advanced Topics in Social Computing

CSCI5070 Advanced Topics in Social Computing CSCI5070 Advanced Topics in Social Computing Irwin King The Chinese University of Hong Kong king@cse.cuhk.edu.hk!! 2012 All Rights Reserved. Outline Scale-Free Networks Generation Properties Analysis Dynamic

More information

Social Data Management Communities

Social Data Management Communities Social Data Management Communities Antoine Amarilli 1, Silviu Maniu 2 January 9th, 2018 1 Télécom ParisTech 2 Université Paris-Sud 1/20 Table of contents Communities in Graphs 2/20 Graph Communities Communities

More information

Phase Transitions in Random Graphs- Outbreak of Epidemics to Network Robustness and fragility

Phase Transitions in Random Graphs- Outbreak of Epidemics to Network Robustness and fragility Phase Transitions in Random Graphs- Outbreak of Epidemics to Network Robustness and fragility Mayukh Nilay Khan May 13, 2010 Abstract Inspired by empirical studies researchers have tried to model various

More information

Comparison of Centralities for Biological Networks

Comparison of Centralities for Biological Networks Comparison of Centralities for Biological Networks Dirk Koschützki and Falk Schreiber Bioinformatics Center Gatersleben-Halle Institute of Plant Genetics and Crop Plant Research Corrensstraße 3 06466 Gatersleben,

More information

A Generating Function Approach to Analyze Random Graphs

A Generating Function Approach to Analyze Random Graphs A Generating Function Approach to Analyze Random Graphs Presented by - Vilas Veeraraghavan Advisor - Dr. Steven Weber Department of Electrical and Computer Engineering Drexel University April 8, 2005 Presentation

More information

Social Network Analysis

Social Network Analysis Social Network Analysis Mathematics of Networks Manar Mohaisen Department of EEC Engineering Adjacency matrix Network types Edge list Adjacency list Graph representation 2 Adjacency matrix Adjacency matrix

More information

Mining Social Network Graphs

Mining Social Network Graphs Mining Social Network Graphs Analysis of Large Graphs: Community Detection Rafael Ferreira da Silva rafsilva@isi.edu http://rafaelsilva.com Note to other teachers and users of these slides: We would be

More information

Overlay (and P2P) Networks

Overlay (and P2P) Networks Overlay (and P2P) Networks Part II Recap (Small World, Erdös Rényi model, Duncan Watts Model) Graph Properties Scale Free Networks Preferential Attachment Evolving Copying Navigation in Small World Samu

More information

Universal Properties of Mythological Networks Midterm report: Math 485

Universal Properties of Mythological Networks Midterm report: Math 485 Universal Properties of Mythological Networks Midterm report: Math 485 Roopa Krishnaswamy, Jamie Fitzgerald, Manuel Villegas, Riqu Huang, and Riley Neal Department of Mathematics, University of Arizona,

More information

(b) Linking and dynamic graph t=

(b) Linking and dynamic graph t= 1 (a) (b) (c) 2 2 2 1 1 1 6 3 4 5 6 3 4 5 6 3 4 5 7 7 7 Supplementary Figure 1: Controlling a directed tree of seven nodes. To control the whole network we need at least 3 driver nodes, which can be either

More information

Modeling web-crawlers on the Internet with random walksdecember on graphs11, / 15

Modeling web-crawlers on the Internet with random walksdecember on graphs11, / 15 Modeling web-crawlers on the Internet with random walks on graphs December 11, 2014 Modeling web-crawlers on the Internet with random walksdecember on graphs11, 2014 1 / 15 Motivation The state of the

More information

Modeling Traffic of Information Packets on Graphs with Complex Topology

Modeling Traffic of Information Packets on Graphs with Complex Topology Modeling Traffic of Information Packets on Graphs with Complex Topology Bosiljka Tadić Jožef Stefan Institute, Box 3000, 1001 Ljubljana, Slovenia Bosiljka.Tadic ijs.si http://phobos.ijs.si/ tadic/ Abstract.

More information

Empirical analysis of online social networks in the age of Web 2.0

Empirical analysis of online social networks in the age of Web 2.0 Physica A 387 (2008) 675 684 www.elsevier.com/locate/physa Empirical analysis of online social networks in the age of Web 2.0 Feng Fu, Lianghuan Liu, Long Wang Center for Systems and Control, College of

More information

Signal Processing for Big Data

Signal Processing for Big Data Signal Processing for Big Data Sergio Barbarossa 1 Summary 1. Networks 2.Algebraic graph theory 3. Random graph models 4. OperaGons on graphs 2 Networks The simplest way to represent the interaction between

More information

TWO CONTRIBUTIONS OF EULER

TWO CONTRIBUTIONS OF EULER TWO CONTRIBUTIONS OF EULER SIEMION FAJTLOWICZ. MATH 4315 Eulerian Tours. Although some mathematical problems which now can be thought of as graph-theoretical, go back to the times of Euclid, the invention

More information

Statistical analysis of the airport network of Pakistan

Statistical analysis of the airport network of Pakistan PRAMANA c Indian Academy of Sciences Vol. 85, No. 1 journal of July 2015 physics pp. 173 183 Statistical analysis of the airport network of Pakistan YASIR TARIQ MOHMAND, AIHU WANG and HAIBIN CHEN School

More information

Finding and Evaluating Fuzzy Clusters in Networks

Finding and Evaluating Fuzzy Clusters in Networks Finding and Evaluating Fuzzy Clusters in Networks Jian Liu LMAM and School of Mathematical Sciences, Peking University, Beijing 100871, P.R. China dugujian@pku.edu.cn Abstract. Fuzzy cluster validity criterion

More information

Module 11. Directed Graphs. Contents

Module 11. Directed Graphs. Contents Module 11 Directed Graphs Contents 11.1 Basic concepts......................... 256 Underlying graph of a digraph................ 257 Out-degrees and in-degrees.................. 258 Isomorphism..........................

More information

networks and the spread of computer viruses

networks and the spread of computer viruses Email networks and the spread of computer viruses M. E. J. Newman, 1 Stephanie Forrest, 1, 2 andjustinbalthrop 2 1 Santa Fe Institute, 1399 Hyde Park Road, Santa Fe, NM 87501 2 Department of Computer Science,

More information

EFFECT OF VARYING THE DELAY DISTRIBUTION IN DIFFERENT CLASSES OF NETWORKS: RANDOM, SCALE-FREE, AND SMALL-WORLD. A Thesis BUM SOON JANG

EFFECT OF VARYING THE DELAY DISTRIBUTION IN DIFFERENT CLASSES OF NETWORKS: RANDOM, SCALE-FREE, AND SMALL-WORLD. A Thesis BUM SOON JANG EFFECT OF VARYING THE DELAY DISTRIBUTION IN DIFFERENT CLASSES OF NETWORKS: RANDOM, SCALE-FREE, AND SMALL-WORLD A Thesis by BUM SOON JANG Submitted to the Office of Graduate Studies of Texas A&M University

More information

Edge-minimal graphs of exponent 2

Edge-minimal graphs of exponent 2 JID:LAA AID:14042 /FLA [m1l; v1.204; Prn:24/02/2017; 12:28] P.1 (1-18) Linear Algebra and its Applications ( ) Contents lists available at ScienceDirect Linear Algebra and its Applications www.elsevier.com/locate/laa

More information

Supplementary Figures

Supplementary Figures Supplementary Figures Cut-off Cut-off Supplementary Figure 1 Setting cut-off to identify links and hidden source. (a) Coarse-grained distribution of element values ln(1 λ i )a ij times 1 in the reconstructed

More information

CSE 258 Lecture 12. Web Mining and Recommender Systems. Social networks

CSE 258 Lecture 12. Web Mining and Recommender Systems. Social networks CSE 258 Lecture 12 Web Mining and Recommender Systems Social networks Social networks We ve already seen networks (a little bit) in week 3 i.e., we ve studied inference problems defined on graphs, and

More information

V2: Measures and Metrics (II)

V2: Measures and Metrics (II) - Betweenness Centrality V2: Measures and Metrics (II) - Groups of Vertices - Transitivity - Reciprocity - Signed Edges and Structural Balance - Similarity - Homophily and Assortative Mixing 1 Betweenness

More information

Maximum number of edges in claw-free graphs whose maximum degree and matching number are bounded

Maximum number of edges in claw-free graphs whose maximum degree and matching number are bounded Maximum number of edges in claw-free graphs whose maximum degree and matching number are bounded Cemil Dibek Tınaz Ekim Pinar Heggernes Abstract We determine the maximum number of edges that a claw-free

More information

Function approximation using RBF network. 10 basis functions and 25 data points.

Function approximation using RBF network. 10 basis functions and 25 data points. 1 Function approximation using RBF network F (x j ) = m 1 w i ϕ( x j t i ) i=1 j = 1... N, m 1 = 10, N = 25 10 basis functions and 25 data points. Basis function centers are plotted with circles and data

More information

Deterministic Hierarchical Networks

Deterministic Hierarchical Networks Deterministic Hierarchical Networks L. Barrière, F. Comellas, C. Dalfó, M.A. Fiol Departament de Matemàtiques Universitat Politècnica de Catalunya Barcelona (Catalonia) {lali.barriere,francesc.comellas}@upc.edu

More information

Integrating local static and dynamic information for routing traffic

Integrating local static and dynamic information for routing traffic Integrating local static and dynamic information for routing traffic Wen-Xu Wang, 1 Chuan-Yang Yin, 1 Gang Yan, 2 and Bing-Hong Wang 1, * 1 Nonlinear Science Center and Department of Modern Physics, University

More information

FITNESS-BASED GENERATIVE MODELS FOR POWER-LAW NETWORKS

FITNESS-BASED GENERATIVE MODELS FOR POWER-LAW NETWORKS Chapter 1 FITNESS-BASED GENERATIVE MODELS FOR POWER-LAW NETWORKS Khanh Nguyen, Duc A. Tran Department of Computer Science University of Massachusets, Boston knguyen,duc@cs.umb.edu Abstract Many real-world

More information

1 Homophily and assortative mixing

1 Homophily and assortative mixing 1 Homophily and assortative mixing Networks, and particularly social networks, often exhibit a property called homophily or assortative mixing, which simply means that the attributes of vertices correlate

More information

ECS 253 / MAE 253, Lecture 8 April 21, Web search and decentralized search on small-world networks

ECS 253 / MAE 253, Lecture 8 April 21, Web search and decentralized search on small-world networks ECS 253 / MAE 253, Lecture 8 April 21, 2016 Web search and decentralized search on small-world networks Search for information Assume some resource of interest is stored at the vertices of a network: Web

More information

CS 534: Computer Vision Segmentation and Perceptual Grouping

CS 534: Computer Vision Segmentation and Perceptual Grouping CS 534: Computer Vision Segmentation and Perceptual Grouping Ahmed Elgammal Dept of Computer Science CS 534 Segmentation - 1 Outlines Mid-level vision What is segmentation Perceptual Grouping Segmentation

More information

The strong chromatic number of a graph

The strong chromatic number of a graph The strong chromatic number of a graph Noga Alon Abstract It is shown that there is an absolute constant c with the following property: For any two graphs G 1 = (V, E 1 ) and G 2 = (V, E 2 ) on the same

More information

FORMATION OF SCATTERNETS WITH HETEROGENEOUS BLUETOOTH DEVICES

FORMATION OF SCATTERNETS WITH HETEROGENEOUS BLUETOOTH DEVICES FORMATION OF SCATTERNETS WITH HETEROGENEOUS BLUETOOTH DEVICES Paal Engelstad, Do Van Thanh, Tore E. Jonvik University of Oslo (UniK) / Telenor R&D, 1331 Fornebu, Norway {Paal.Engelstad, Thanh-van.Do, Tore-erling.Jonvik

More information

The Topology and Dynamics of Complex Man- Made Systems

The Topology and Dynamics of Complex Man- Made Systems The Topology and Dynamics of Complex Man- Made Systems Dan Braha New England Complex Institute Cambridge, MA, USA University of Massachusetts Dartmouth, MA, USA braha@necsi.edu http://necsi.edu/affiliates/braha/dan_braha-description.htm

More information

An introduction to the physics of complex networks

An introduction to the physics of complex networks An introduction to the physics of complex networks Alain Barrat CPT, Marseille, France ISI, Turin, Italy http://www.cpt.univ-mrs.fr/~barrat http://www.cxnets.org http://www.sociopatterns.org REVIEWS: Statistical

More information

The Buss Reduction for the k-weighted Vertex Cover Problem

The Buss Reduction for the k-weighted Vertex Cover Problem The Buss Reduction for the k-weighted Vertex Cover Problem Hong Xu Xin-Zeng Wu Cheng Cheng Sven Koenig T. K. Satish Kumar University of Southern California, Los Angeles, California 90089, the United States

More information

From Centrality to Temporary Fame: Dynamic Centrality in Complex Networks

From Centrality to Temporary Fame: Dynamic Centrality in Complex Networks From Centrality to Temporary Fame: Dynamic Centrality in Complex Networks Dan Braha 1, 2 and Yaneer Bar-Yam 2 1 University of Massachusetts Dartmouth, MA 02747, USA 2 New England Complex Systems Institute

More information

Clustering. SC4/SM4 Data Mining and Machine Learning, Hilary Term 2017 Dino Sejdinovic

Clustering. SC4/SM4 Data Mining and Machine Learning, Hilary Term 2017 Dino Sejdinovic Clustering SC4/SM4 Data Mining and Machine Learning, Hilary Term 2017 Dino Sejdinovic Clustering is one of the fundamental and ubiquitous tasks in exploratory data analysis a first intuition about the

More information

degree at least en? Unfortunately, we can throw very little light on this simple question. Our only result in this direction (Theorem 3) is that, if w

degree at least en? Unfortunately, we can throw very little light on this simple question. Our only result in this direction (Theorem 3) is that, if w REMARKS ON STARS AND INDEPENDENT SETS P. Erdös and J. Pach Mathematical Institute of the Hungarian Academy of Sciences 1 INTRODUCTION Let G be a graph with vertex set and edge set V(G) and E(G), respectively.

More information

Eciency of scale-free networks: error and attack tolerance

Eciency of scale-free networks: error and attack tolerance Available online at www.sciencedirect.com Physica A 320 (2003) 622 642 www.elsevier.com/locate/physa Eciency of scale-free networks: error and attack tolerance Paolo Crucitti a, Vito Latora b, Massimo

More information

Machine Learning and Modeling for Social Networks

Machine Learning and Modeling for Social Networks Machine Learning and Modeling for Social Networks Olivia Woolley Meza, Izabela Moise, Nino Antulov-Fatulin, Lloyd Sanders 1 Introduction to Networks Computational Social Science D-GESS Olivia Woolley Meza

More information

Edge-exchangeable graphs and sparsity

Edge-exchangeable graphs and sparsity Edge-exchangeable graphs and sparsity Tamara Broderick Department of EECS Massachusetts Institute of Technology tbroderick@csail.mit.edu Diana Cai Department of Statistics University of Chicago dcai@uchicago.edu

More information

arxiv: v1 [physics.soc-ph] 13 Jan 2011

arxiv: v1 [physics.soc-ph] 13 Jan 2011 An Estimation of the Shortest and Largest Average Path Length in Graphs of Given Density László Gulyás, Gábor Horváth, Tamás Cséri and George Kampis arxiv:1101.549v1 [physics.soc-ph] 1 Jan 011 Abstract

More information

1. Introduction. 2. Motivation and Problem Definition. Volume 8 Issue 2, February Susmita Mohapatra

1. Introduction. 2. Motivation and Problem Definition. Volume 8 Issue 2, February Susmita Mohapatra Pattern Recall Analysis of the Hopfield Neural Network with a Genetic Algorithm Susmita Mohapatra Department of Computer Science, Utkal University, India Abstract: This paper is focused on the implementation

More information

Immunization for complex network based on the effective degree of vertex

Immunization for complex network based on the effective degree of vertex Immunization for complex network based on the effective degree of vertex Ke Hu and Yi Tang * Department of Physics and Institute of Modern Physics, Xiangtan University, Xiangtan 411105, Hunan, China The

More information

Discovery of Community Structure in Complex Networks Based on Resistance Distance and Center Nodes

Discovery of Community Structure in Complex Networks Based on Resistance Distance and Center Nodes Journal of Computational Information Systems 8: 23 (2012) 9807 9814 Available at http://www.jofcis.com Discovery of Community Structure in Complex Networks Based on Resistance Distance and Center Nodes

More information

Incoming, Outgoing Degree and Importance Analysis of Network Motifs

Incoming, Outgoing Degree and Importance Analysis of Network Motifs Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology IJCSMC, Vol. 4, Issue. 6, June 2015, pg.758

More information

Simplicial Complexes of Networks and Their Statistical Properties

Simplicial Complexes of Networks and Their Statistical Properties Simplicial Complexes of Networks and Their Statistical Properties Slobodan Maletić, Milan Rajković*, and Danijela Vasiljević Institute of Nuclear Sciences Vinča, elgrade, Serbia *milanr@vin.bg.ac.yu bstract.

More information

c 2004 Society for Industrial and Applied Mathematics

c 2004 Society for Industrial and Applied Mathematics SIAM J. MATRIX ANAL. APPL. Vol. 26, No. 2, pp. 390 399 c 2004 Society for Industrial and Applied Mathematics HERMITIAN MATRICES, EIGENVALUE MULTIPLICITIES, AND EIGENVECTOR COMPONENTS CHARLES R. JOHNSON

More information

CAIM: Cerca i Anàlisi d Informació Massiva

CAIM: Cerca i Anàlisi d Informació Massiva 1 / 72 CAIM: Cerca i Anàlisi d Informació Massiva FIB, Grau en Enginyeria Informàtica Slides by Marta Arias, José Balcázar, Ricard Gavaldá Department of Computer Science, UPC Fall 2016 http://www.cs.upc.edu/~caim

More information

Maximal Monochromatic Geodesics in an Antipodal Coloring of Hypercube

Maximal Monochromatic Geodesics in an Antipodal Coloring of Hypercube Maximal Monochromatic Geodesics in an Antipodal Coloring of Hypercube Kavish Gandhi April 4, 2015 Abstract A geodesic in the hypercube is the shortest possible path between two vertices. Leader and Long

More information