Skip to main content
arXiv is now an independent nonprofit! Learn more

Showing 1–40 of 40 results for author: Galstyan, A

Searching in archive physics. Search in all archives.
.
  1. arXiv:2111.02434  [pdf, other

    cs.LG physics.comp-ph

    Hamiltonian Dynamics with Non-Newtonian Momentum for Rapid Sampling

    Authors: Greg Ver Steeg, Aram Galstyan

    Abstract: Sampling from an unnormalized probability distribution is a fundamental problem in machine learning with applications including Bayesian modeling, latent factor inference, and energy-based model training. After decades of research, variations of MCMC remain the default approach to sampling despite slow convergence. Auxiliary neural models can learn to speed up MCMC, but the overhead for training t… ▽ More

    Submitted 29 December, 2021; v1 submitted 3 November, 2021; originally announced November 2021.

    Comments: 31 pages, 19 figures. Advances in Neural Information Processing Systems (NeurIPS), 2021. Animations at https://sites.google.com/view/esh-dynamics/home, code at https://github.com/gregversteeg/esh_dynamics

  2. arXiv:1909.07578  [pdf, other

    stat.ML cs.LG cs.SI physics.data-an q-bio.MN

    Stacking Models for Nearly Optimal Link Prediction in Complex Networks

    Authors: Amir Ghasemian, Homa Hosseinmardi, Aram Galstyan, Edoardo M. Airoldi, Aaron Clauset

    Abstract: Most real-world networks are incompletely observed. Algorithms that can accurately predict which links are missing can dramatically speedup the collection of network data and improve the validity of network models. Many algorithms now exist for predicting missing links, given a partially observed network, but it has remained unknown whether a single best predictor exists, how link predictability v… ▽ More

    Submitted 17 September, 2019; originally announced September 2019.

    Comments: 30 pages, 9 figures, 22 tables

    Journal ref: Proc. Natl. Acad. Sci. USA 117(38), 23393-23400 (2020)

  3. arXiv:1903.08136  [pdf, other

    cs.SI physics.soc-ph

    Debiasing Community Detection: The Importance of Lowly-Connected Nodes

    Authors: Ninareh Mehrabi, Fred Morstatter, Nanyun Peng, Aram Galstyan

    Abstract: Community detection is an important task in social network analysis, allowing us to identify and understand the communities within the social structures. However, many community detection approaches either fail to assign low degree (or lowly-connected) users to communities, or assign them to trivially small communities that prevent them from being included in analysis. In this work, we investigate… ▽ More

    Submitted 19 March, 2019; originally announced March 2019.

  4. Coherent Compton scattering from hydrogen and helium atoms

    Authors: Irina A. Gnilozub, Alexander Galstyan, Yuri V. Popov, Igor P. Volobuev

    Abstract: We develop an approach to describing coherent Compton scattering of photons in the keV energy range from hydrogen and helium atoms based on a relativistic version of the AA-approximation within the standard perturbative S-matrix formalism. The resulting formulas for the cross section take into account the effects of the electron boundness and correctly reproduce the behavior of the cross section a… ▽ More

    Submitted 24 September, 2018; originally announced September 2018.

    Comments: 17 pages, 5 figures

  5. arXiv:1803.06638  [pdf, ps, other

    physics.data-an cond-mat.stat-mech cs.GT

    Adaptive Decision Making via Entropy Minimization

    Authors: Armen E. Allahverdyan, Aram Galstyan, Ali E. Abbas, Zbigniew R. Struzik

    Abstract: An agent choosing between various actions tends to take the one with the lowest cost. But this choice is arguably too rigid (not adaptive) to be useful in complex situations, e.g., where exploration-exploitation trade-off is relevant in creative task solving or when stated preferences differ from revealed ones. Here we study an agent who is willing to sacrifice a fixed amount of expected utility f… ▽ More

    Submitted 1 December, 2018; v1 submitted 18 March, 2018; originally announced March 2018.

    Comments: 21 pages, 3 figures

    Journal ref: International Journal of Approximate Reasoning, 103, 270-287 (2018)

  6. arXiv:1712.07844  [pdf, ps, other

    physics.soc-ph cond-mat.dis-nn cs.SI

    Memory-induced mechanism for self-sustaining activity in networks

    Authors: A. E. Allahverdyan, G. Ver Steeg, A. Galstyan

    Abstract: We study a mechanism of activity sustaining on networks inspired by a well-known model of neuronal dynamics. Our primary focus is the emergence of self-sustaining collective activity patterns, where no single node can stay active by itself, but the activity provided initially is sustained within the collective of interacting agents. In contrast to existing models of self-sustaining activity that a… ▽ More

    Submitted 21 December, 2017; originally announced December 2017.

    Comments: 23 pages, 12 figures

    Journal ref: Physical Review E 92, 062824 (2015)

  7. arXiv:1711.11350  [pdf, other

    physics.atom-ph

    Static field limit of excitation probabilities in laser-atom interactions

    Authors: A. Galstyan, V. L. Shablov, Yu. V. Popov, F. Mota-Furtado, P. F. O'Mahony, B. Piraux

    Abstract: We consider the interaction of atomic hydrogen, in its ground state, with an electromagnetic pulse whose duration is fixed in terms of the number of optical cycles. We study the probability of excitation of the atom in the static field limit i.e. for field frequencies going to zero. Despite the fact that the well known Born-Fock adiabatic theorem is valid only for a system whose energy spectrum is… ▽ More

    Submitted 14 December, 2018; v1 submitted 30 November, 2017; originally announced November 2017.

  8. arXiv:1710.09076  [pdf, ps, other

    physics.soc-ph cond-mat.dis-nn cs.SI

    Emergence of Leadership in Communication

    Authors: Armen E. Allahverdyan, Aram Galstyan

    Abstract: We study a neuro-inspired model that mimics a discussion (or information dissemination) process in a network of agents. During their interaction, agents redistribute activity and network weights, resulting in emergence of leader(s). The model is able to reproduce the basic scenarios of leadership known in nature and society: laissez-faire (irregular activity, weak leadership, sizable inter-followe… ▽ More

    Submitted 25 October, 2017; originally announced October 2017.

    Comments: 17 pages, 11 figures

    Journal ref: PLoS ONE 11, e0159301 (2016)

  9. Fully differential cross sections for singly ionizing 1-Mev p+He collisions at small momentum transfer: Beyond the first Born approximation

    Authors: O. Chuluunbaatar, S. A. Zaytsev, K. A. Kouzakov, A. Galstyan, V. L. Shablov, Yu. V. Popov

    Abstract: We present calculations of the electron angular distributions in the single ionization of helium by 1-MeV proton impact at momentum transfer of 0.75 a.u. and ejected-electron energy of 6.5 eV. The results using the first and second Born approximations and the 3C model with different trial helium functions are compared to the experimental data. A good agreement between theory and experiment is foun… ▽ More

    Submitted 1 November, 2017; v1 submitted 15 October, 2017; originally announced October 2017.

    Comments: 19 pages, 5 figures

    Journal ref: Physical Review A 96, 042716 (2017)

  10. Ionisation of H$_2$O by a strong ultrashort XUV pulse: a model within the single active electron approximation

    Authors: A. Galstyan, Yu. V. Popov, N. Janssens, F. Mota-Furtado, P. F. O'Mahony, P. Decleva, N. Quadri, B. Piraux

    Abstract: We present and discuss a new computationally inexpensive method to study, within the single active electron approximation, the interaction of a complex system with an intense ultrashort laser pulse. As a first application, we consider the one photon single ionisation of the highest occupied molecular orbital of the water molecule by a laser pulse. The ionisation yield is calculated for different o… ▽ More

    Submitted 4 September, 2017; v1 submitted 16 March, 2017; originally announced March 2017.

    Comments: 24 pages, 13 figures

  11. The strong field approximation within a Faddeev-like formalism for laser-matter interactions

    Authors: Yu. Popov, A. Galstyan, F. Mota-Furtado, P. F. O'Mahony, B. Piraux

    Abstract: We consider the interaction of atomic hydrogen with an intense laser field within the strong-field approximation. By using a Faddeev-like formalism, we introduce a new perturbative series in the binding potential of the atom. As a first test of this new approach, we calculate the electron energy spectrum in the very simple case of a photon energy higher than the ionisation potential. We show that… ▽ More

    Submitted 14 November, 2016; originally announced November 2016.

    Comments: 7 pages, 1 figure

  12. Modelling laser-atom interactions in the strong field regime

    Authors: A. Galstyan, Yu. V. Popov, F. Mota-Furtado, P. F. O'Mahony, N. Janssens, S. D. Jenkins, O. Chuluunbaatar, B. Piraux

    Abstract: We consider the ionisation of atomic hydrogen by a strong infrared field. We extend and study in more depth an existing semi-analytical model. Starting from the time-dependent Schroedinger equation in momentum space and in the velocity gauge we substitute the kernel of the non-local Coulomb potential by a sum of N separable potentials, each of them supporting one hydrogen bound state. This leads t… ▽ More

    Submitted 14 November, 2016; originally announced November 2016.

    Comments: 11 pages, 5 figures

  13. arXiv:1605.00659  [pdf, other

    cs.SI cs.LG physics.soc-ph

    Predicting online extremism, content adopters, and interaction reciprocity

    Authors: Emilio Ferrara, Wen-Qiang Wang, Onur Varol, Alessandro Flammini, Aram Galstyan

    Abstract: We present a machine learning framework that leverages a mixture of metadata, network, and temporal features to detect extremist users, and predict content adopters and interaction reciprocity in social media. We exploit a unique dataset containing millions of tweets generated by more than 25 thousand users who have been manually identified, reported, and suspended by Twitter due to their involvem… ▽ More

    Submitted 2 May, 2016; originally announced May 2016.

    Comments: 9 pages, 3 figures, 8 tables

    Journal ref: International Conference on Social Informatics (pp. 22-39). Springer. 2016

  14. arXiv:1601.05140  [pdf

    cs.SI cs.AI cs.CY physics.data-an physics.soc-ph

    The DARPA Twitter Bot Challenge

    Authors: V. S. Subrahmanian, Amos Azaria, Skylar Durst, Vadim Kagan, Aram Galstyan, Kristina Lerman, Linhong Zhu, Emilio Ferrara, Alessandro Flammini, Filippo Menczer, Andrew Stevens, Alexander Dekhtyar, Shuyang Gao, Tad Hogg, Farshad Kooti, Yan Liu, Onur Varol, Prashant Shiralkar, Vinod Vydiswaran, Qiaozhu Mei, Tim Hwang

    Abstract: A number of organizations ranging from terrorist groups such as ISIS to politicians and nation states reportedly conduct explicit campaigns to influence opinion on social media, posing a risk to democratic processes. There is thus a growing need to identify and eliminate "influence bots" - realistic, automated identities that illicitly shape discussion on sites like Twitter and Facebook - before t… ▽ More

    Submitted 21 April, 2016; v1 submitted 19 January, 2016; originally announced January 2016.

    Comments: IEEE Computer Magazine, in press

    Journal ref: Computer 49 (6), 38-46. IEEE, 2016

  15. arXiv:1512.00681  [pdf, other

    physics.atom-ph quant-ph

    Reformulation of the strong field approximation for light-matter interactions

    Authors: A. Galstyan, O. Chuluunbaatar, A. Hamido, Yu. V. Popov, F. Mota-Furtado, P. F. O'Mahony, N. Janssens, F. Catoire, B. Piraux

    Abstract: We consider the interaction of hydrogen-like atoms with a strong laser field and show that the strong field approximation and all its variants may be grouped into a set of families of approximation schemes. This is done by introducing an ansatz describing the electron wave packet as the sum of the initial state wave function times a phase factor and a function which is the perturbative solution in… ▽ More

    Submitted 2 December, 2015; originally announced December 2015.

    Comments: 19 pages, 9 figures, submitted for publication

    Journal ref: Phys. Rev. A 93, 023422 (2016)

  16. arXiv:1510.05318  [pdf, other

    cs.SI cs.LG physics.data-an physics.soc-ph

    Latent Space Model for Multi-Modal Social Data

    Authors: Yoon-Sik Cho, Greg Ver Steeg, Emilio Ferrara, Aram Galstyan

    Abstract: With the emergence of social networking services, researchers enjoy the increasing availability of large-scale heterogenous datasets capturing online user interactions and behaviors. Traditional analysis of techno-social systems data has focused mainly on describing either the dynamics of social interactions, or the attributes and behaviors of the users. However, overwhelming empirical evidence su… ▽ More

    Submitted 18 October, 2015; originally announced October 2015.

    Comments: 12 pages, 7 figures, 2 tables

    Journal ref: Proceedings of the 25th International Conference on World Wide Web (pp. 447-458). 2016

  17. arXiv:1508.00536  [pdf, other

    cs.IT physics.data-an

    Estimating Mutual Information by Local Gaussian Approximation

    Authors: Shuyang Gao, Greg Ver Steeg, Aram Galstyan

    Abstract: Estimating mutual information (MI) from samples is a fundamental problem in statistics, machine learning, and data analysis. Recently it was shown that a popular class of non-parametric MI estimators perform very poorly for strongly dependent variables and have sample complexity that scales exponentially with the true MI. This undesired behavior was attributed to the reliance of those estimators o… ▽ More

    Submitted 17 February, 2016; v1 submitted 3 August, 2015; originally announced August 2015.

    Comments: To appear in the 31st Conference on Uncertainty in Artificial Intelligence (UAI), 2015

  18. arXiv:1412.0696  [pdf, other

    cs.CL cs.IT cs.SI physics.data-an

    Understanding confounding effects in linguistic coordination: an information-theoretic approach

    Authors: Shuyang Gao, Greg Ver Steeg, Aram Galstyan

    Abstract: We suggest an information-theoretic approach for measuring stylistic coordination in dialogues. The proposed measure has a simple predictive interpretation and can account for various confounding factors through proper conditioning. We revisit some of the previous studies that reported strong signatures of stylistic accommodation, and find that a significant part of the observed coordination can b… ▽ More

    Submitted 27 August, 2015; v1 submitted 1 December, 2014; originally announced December 2014.

    Journal ref: PLoS ONE 10(6): e0130167, 2015

  19. arXiv:1411.4328  [pdf, ps, other

    physics.soc-ph cs.SI

    Opinion Dynamics with Confirmation Bias

    Authors: A. E. Allahverdyan, Aram Galstyan

    Abstract: Background: Confirmation bias is the tendency to acquire or evaluate new information in a way that is consistent with one's preexisting beliefs. It is omnipresent in psychology, economics, and even scientific practices. Prior theoretical research of this phenomenon has mainly focused on its economic implications possibly missing its potential connections with broader notions of cognitive science.… ▽ More

    Submitted 16 November, 2014; originally announced November 2014.

    Comments: 18 pages

    Journal ref: PLoS ONE 9(7), e99557 (2014)

  20. arXiv:1411.2003  [pdf, other

    cs.IT physics.data-an stat.ML

    Efficient Estimation of Mutual Information for Strongly Dependent Variables

    Authors: Shuyang Gao, Greg Ver Steeg, Aram Galstyan

    Abstract: We demonstrate that a popular class of nonparametric mutual information (MI) estimators based on k-nearest-neighbor graphs requires number of samples that scales exponentially with the true MI. Consequently, accurate estimation of MI between two strongly dependent variables is possible only for prohibitively large sample size. This important yet overlooked shortcoming of the existing estimators is… ▽ More

    Submitted 5 March, 2015; v1 submitted 7 November, 2014; originally announced November 2014.

    Comments: 13 pages, to appear in International Conference on Artificial Intelligence and Statistics (AISTATS) 2015

  21. arXiv:1410.7404  [pdf, other

    stat.ML cs.LG physics.data-an

    Maximally Informative Hierarchical Representations of High-Dimensional Data

    Authors: Greg Ver Steeg, Aram Galstyan

    Abstract: We consider a set of probabilistic functions of some input variables as a representation of the inputs. We present bounds on how informative a representation is about input data. We extend these bounds to hierarchical representations so that we can quantify the contribution of each layer towards capturing the information in the original data. The special form of these bounds leads to a simple, bot… ▽ More

    Submitted 30 January, 2015; v1 submitted 27 October, 2014; originally announced October 2014.

    Comments: 13 pages, 8 figures. Appearing in Proceedings of the 18th International Conference on Artificial Intelligence and Statistics (AISTATS) 2015

  22. arXiv:1312.0631  [pdf, ps, other

    cs.SI cond-mat.stat-mech physics.soc-ph stat.ML

    Phase Transitions in Community Detection: A Solvable Toy Model

    Authors: Greg Ver Steeg, Cristopher Moore, Aram Galstyan, Armen E. Allahverdyan

    Abstract: Recently, it was shown that there is a phase transition in the community detection problem. This transition was first computed using the cavity method, and has been proved rigorously in the case of $q=2$ groups. However, analytic calculations using the cavity method are challenging since they require us to understand probability distributions of messages. We study analogous transitions in so-calle… ▽ More

    Submitted 2 December, 2013; originally announced December 2013.

    Comments: 6 pages, 6 figures

  23. Transfer excitation reactions in fast proton-helium collisions

    Authors: M. S. Schöffler, H. -K. Kim, O. Chuluunbaatar, S. Houamer, A. G. Galstyan, J. N. Titze, T. Jahnke, L. Ph. H. Schmidt, H. Schmidt-B"ocking, R. D"orner, Yu. V. Popov, A. A. Bulychev

    Abstract: Continuing previous work, we have measured the projectile scattering-angle dependency for transfer excitation of fast protons (300-1200 keV/u) colliding with helium (p+He $\rightarrow$ H + He$^{+ *}$). Our high-resolution fully differential data are accompanied by calculations, performed in the plane wave first Born approximation and the eikonal wave Born approximation. Experimentally we find a de… ▽ More

    Submitted 11 March, 2014; v1 submitted 22 November, 2013; originally announced November 2013.

    Comments: published in Phys. Rev. A (2014)

    Journal ref: Phys. Rev. A 89, 032707 (2014)

  24. arXiv:1310.4210  [pdf, other

    cs.LG cs.IT physics.data-an stat.ML

    Demystifying Information-Theoretic Clustering

    Authors: Greg Ver Steeg, Aram Galstyan, Fei Sha, Simon DeDeo

    Abstract: We propose a novel method for clustering data which is grounded in information-theoretic principles and requires no parametric assumptions. Previous attempts to use information theory to define clusters in an assumption-free way are based on maximizing mutual information between data and cluster labels. We demonstrate that this intuition suffers from a fundamental conceptual flaw that causes clust… ▽ More

    Submitted 5 February, 2014; v1 submitted 15 October, 2013; originally announced October 2013.

    Comments: Proceedings of The 31st International Conference on Machine Learning (ICML), 2014. 11 pages, 9 figures

  25. 2D electron momentum distributions for transfer ionization in fast proton Helium collisions

    Authors: M. S. Schoeffler, O. Chuluunbaatar, S. Houamer, A. G. Galstyan, J. N. Titze, L. Ph. H. Schmidt, T. Jahnke, H. Schmidt-Boecking, R. Doerner, Yu. V. Popov, A. A. Gusev, C. Dal Cappello

    Abstract: The momentum distribution of the electron in the reaction p+He $\rightarrow$ H + He$^{2+}$ + $e$ is measured for projectile energies $E_p$=300 and 630 keV/u at very small scattering angles of hydrogen. We mainly present two dimensional distributions parallel $(k_{||})$ and perpendicular $(k_{\perp})$ to the projectile beam. Theoretical calculations were carried out within the Plane Wave First Born… ▽ More

    Submitted 19 May, 2013; originally announced May 2013.

    Comments: 14 pages, 4 figures

  26. arXiv:1305.1980  [pdf, other

    physics.soc-ph cs.SI physics.data-an stat.AP

    Modeling Temporal Activity Patterns in Dynamic Social Networks

    Authors: Vasanthan Raghavan, Greg Ver Steeg, Aram Galstyan, Alexander G. Tartakovsky

    Abstract: The focus of this work is on developing probabilistic models for user activity in social networks by incorporating the social network influence as perceived by the user. For this, we propose a coupled Hidden Markov Model, where each user's activity evolves according to a Markov chain with a hidden state that is influenced by the collective activity of the friends of the user. We develop generalize… ▽ More

    Submitted 8 May, 2013; originally announced May 2013.

    Comments: 23 pages, 7 figures, 13 tables, submitted for publication

  27. Comment on "Dynamics of transfer ionization in fast ion-atom collisions"

    Authors: Yu. V. Popov, V. L. Shablov, K. A. Kouzakov, A. G. Galstyan

    Abstract: We inspect the first-order electron-electron capture scenario for transfer ionization that has been recently formulated by Voitkiv et al. (Phys. Rev. A 86, 012709 (2012) and references therein). Using the multichannel scattering theory for many-body systems with Coulomb interactions, we show that this scenario is just a part of the well-studied Oppenheimer-Brinkmann-Kramers approximation. Accurate… ▽ More

    Submitted 11 August, 2013; v1 submitted 10 April, 2013; originally announced April 2013.

    Comments: 1 figure

  28. arXiv:1211.4889  [pdf, other

    cs.SI physics.soc-ph stat.ME

    Statistical Tests for Contagion in Observational Social Network Studies

    Authors: Greg Ver Steeg, Aram Galstyan

    Abstract: Current tests for contagion in social network studies are vulnerable to the confounding effects of latent homophily (i.e., ties form preferentially between individuals with similar hidden traits). We demonstrate a general method to lower bound the strength of causal effects in observational social network studies, even in the presence of arbitrary, unobserved individual traits. Our tests require n… ▽ More

    Submitted 15 April, 2013; v1 submitted 20 November, 2012; originally announced November 2012.

    Comments: 9 pages, 4 figures. Appearing at AISTATS-13

  29. arXiv:1208.4475  [pdf, other

    cs.SI physics.soc-ph stat.AP

    Information-Theoretic Measures of Influence Based on Content Dynamics

    Authors: Greg Ver Steeg, Aram Galstyan

    Abstract: The fundamental building block of social influence is for one person to elicit a response in another. Researchers measuring a "response" in social media typically depend either on detailed models of human behavior or on platform-specific cues such as re-tweets, hash tags, URLs, or mentions. Most content on social networks is difficult to model because the modes and motivation of human expression a… ▽ More

    Submitted 15 February, 2013; v1 submitted 22 August, 2012; originally announced August 2012.

    Comments: 10 pages, 10 figures. In Proceedings of the 6th International Conference on Web Search and Data Mining. (v2: title change, v3: minor edits)

  30. Transfer ionization and its sensitivity to the ground-state wave function

    Authors: M. S. Schöffler, O. Chuluunbaatar, Yu. V. Popov, S. Houamer, J. Titze, T. Jahnke, L. Ph. H. Schmidt, O. Jagutzki, A. G. Galstyan, A. A. Gusev

    Abstract: We present kinematically complete theoretical calculations and experiments for transfer ionization in H$^++$He collisions at 630 keV/u. Experiment and theory are compared on the most detailed level of fully differential cross sections in the momentum space. This allows us to unambiguously identify contributions from the shake-off and two-step-2 mechanisms of the reaction. It is shown that the simu… ▽ More

    Submitted 22 February, 2013; v1 submitted 6 August, 2012; originally announced August 2012.

    Journal ref: Phys. Rev. A 87, 032715 (2013)

  31. arXiv:1207.1497  [pdf, ps, other

    stat.AP cs.SI physics.data-an physics.soc-ph

    Hidden Markov models for the activity profile of terrorist groups

    Authors: Vasanthan Raghavan, Aram Galstyan, Alexander G. Tartakovsky

    Abstract: The main focus of this work is on developing models for the activity profile of a terrorist group, detecting sudden spurts and downfalls in this profile, and, in general, tracking it over a period of time. Toward this goal, a $d$-state hidden Markov model (HMM) that captures the latent states underlying the dynamics of the group and thus its activity profile is developed. The simplest setting of… ▽ More

    Submitted 15 January, 2014; v1 submitted 5 July, 2012; originally announced July 2012.

    Comments: Published in at http://dx.doi.org/10.1214/13-AOAS682 the Annals of Applied Statistics (http://www.imstat.org/aoas/) by the Institute of Mathematical Statistics (http://www.imstat.org)

    Report number: IMS-AOAS-AOAS682

    Journal ref: Annals of Applied Statistics 2013, Vol. 7, No. 4, 2402-2430

  32. Effects of nonzero photon momentum in (γ,2e) processes

    Authors: Alexander G. Galstyan, Ochbadrakh Chuluunbaatar, Yuri V. Popov, Bernard Piraux

    Abstract: We study the effects of nonzero photon momentum on the triply-differential cross section for (γ,2e) processes. Due to the low value of the photon momentum, these effects are weak and manifest only in special kinematical conditions like the back-to-back emission of the electrons with equal energy sharing. Helium and a few light helium-like ions are treated in detail. Quite unexpectedly, the magnitu… ▽ More

    Submitted 5 January, 2012; originally announced January 2012.

    Comments: 8 pages, 7 figures, 1 table

  33. arXiv:1110.2724  [pdf, other

    cs.SI physics.soc-ph stat.AP

    Information Transfer in Social Media

    Authors: Greg Ver Steeg, Aram Galstyan

    Abstract: Recent research has explored the increasingly important role of social media by examining the dynamics of individual and group behavior, characterizing patterns of information diffusion, and identifying influential individuals. In this paper we suggest a measure of causal relationships between nodes based on the information-theoretic notion of transfer entropy, or information transfer. This theore… ▽ More

    Submitted 12 October, 2011; originally announced October 2011.

    Comments: 8 pages, 10 figures

  34. arXiv:1106.2788  [pdf, other

    cs.SI physics.soc-ph stat.ML

    Co-evolution of Selection and Influence in Social Networks

    Authors: Yoon-Sik Cho, Greg Ver Steeg, Aram Galstyan

    Abstract: Many networks are complex dynamical systems, where both attributes of nodes and topology of the network (link structure) can change with time. We propose a model of co-evolving networks where both node at- tributes and network structure evolve under mutual influence. Specifically, we consider a mixed membership stochastic blockmodel, where the probability of observing a link between two nodes depe… ▽ More

    Submitted 14 June, 2011; originally announced June 2011.

    Comments: In Proc. of the Twenty-Fifth Conference on Artificial Intelligence (AAAI-11)

  35. arXiv:1106.1636  [pdf, other

    cs.AI cs.SI physics.soc-ph quant-ph stat.ML

    A Sequence of Relaxations Constraining Hidden Variable Models

    Authors: Greg Ver Steeg, Aram Galstyan

    Abstract: Many widely studied graphical models with latent variables lead to nontrivial constraints on the distribution of the observed variables. Inspired by the Bell inequalities in quantum mechanics, we refer to any linear inequality whose violation rules out some latent variable model as a "hidden variable test" for that model. Our main contribution is to introduce a sequence of relaxations which provid… ▽ More

    Submitted 20 July, 2011; v1 submitted 8 June, 2011; originally announced June 2011.

    Comments: UAI 2011 Best Paper Runner-Up; Proceedings of the 27th Conference on Uncertainty in Artificial Intelligence (UAI 2011)

  36. arXiv:1101.4227  [pdf, ps, other

    physics.data-an cond-mat.dis-nn cond-mat.stat-mech cs.LG

    Statistical Mechanics of Semi-Supervised Clustering in Sparse Graphs

    Authors: Greg Ver Steeg, Aram Galstyan, Armen E. Allahverdyan

    Abstract: We theoretically study semi-supervised clustering in sparse graphs in the presence of pairwise constraints on the cluster assignments of nodes. We focus on bi-cluster graphs, and study the impact of semi-supervision for varying constraint density and overlap between the clusters. Recent results for unsupervised clustering in sparse graphs indicate that there is a critical ratio of within-cluster a… ▽ More

    Submitted 30 October, 2011; v1 submitted 21 January, 2011; originally announced January 2011.

    Comments: 8 pages, 4 figures

    Journal ref: J. Stat. Mech. (2011) P08009

  37. arXiv:0907.4803  [pdf, ps, other

    physics.soc-ph cond-mat.stat-mech cs.CY physics.data-an q-bio.QM

    Community Detection with and without Prior Information

    Authors: Armen E. Allahverdyan, Greg Ver Steeg, Aram Galstyan

    Abstract: We study the problem of graph partitioning, or clustering, in sparse networks with prior information about the clusters. Specifically, we assume that for a fraction $ρ$ of the nodes their true cluster assignments are known in advance. This can be understood as a semi--supervised version of clustering, in contrast to unsupervised clustering where the only available information is the graph structur… ▽ More

    Submitted 5 October, 2010; v1 submitted 27 July, 2009; originally announced July 2009.

    Comments: 6 pages, 2 figures

    Journal ref: Europhysics Letters 90, p.18002, 2010

  38. arXiv:0906.1980  [pdf, ps, other

    cs.AI cond-mat.stat-mech cs.IT physics.data-an stat.ML

    On Maximum a Posteriori Estimation of Hidden Markov Processes

    Authors: Armen Allahverdyan, Aram Galstyan

    Abstract: We present a theoretical analysis of Maximum a Posteriori (MAP) sequence estimation for binary symmetric hidden Markov processes. We reduce the MAP estimation to the energy minimization of an appropriately defined Ising spin model, and focus on the performance of MAP as characterized by its accuracy and the number of solutions corresponding to a typical observed sequence. It is shown that for a… ▽ More

    Submitted 10 June, 2009; originally announced June 2009.

    Comments: 9 pages, to appear in the Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence (UAI-2009)

  39. arXiv:0905.1108  [pdf, other

    physics.soc-ph cond-mat.stat-mech cs.CY

    Maximizing Influence Propagation in Networks with Community Structure

    Authors: Aram Galstyan, Vahe Musoyan, Paul Cohen

    Abstract: We consider the algorithmic problem of selecting a set of target nodes that cause the biggest activation cascade in a network. In case when the activation process obeys the diminishing returns property, a simple hill-climbing selection mechanism has been shown to achieve a provably good performance. Here we study models of influence propagation that exhibit critical behavior, and where the prope… ▽ More

    Submitted 7 May, 2009; originally announced May 2009.

    Comments: 7 pages, 8 figures

    Journal ref: Phys. Rev. E 79, 056102 (2009)

  40. arXiv:0803.1932  [pdf, other

    cond-mat.stat-mech cond-mat.dis-nn physics.soc-ph

    Cascading Dynamics in Modular Networks

    Authors: Aram Galstyan, Paul Cohen

    Abstract: In this paper we study a simple cascading process in a structured heterogeneous population, namely, a network composed of two loosely coupled communities. We demonstrate that under certain conditions the cascading dynamics in such a network has a two--tiered structure that characterizes activity spreading at different rates in the communities. We study the dynamics of the model using both simula… ▽ More

    Submitted 13 March, 2008; originally announced March 2008.

    Comments: 5 pages, 4 figures

    Journal ref: Phys. Rev. E 75, 036109 (2007)