Bulletin of the American Physical Society
APS March Meeting 2014
Volume 59, Number 1
Monday–Friday, March 3–7, 2014; Denver, Colorado
Session B16: Statistical Mechanics of Social Systems |
Hide Abstracts |
Sponsoring Units: GSNP Chair: G Korniss Room: 401 |
Monday, March 3, 2014 11:15AM - 11:27AM |
B16.00001: Statistical Mechanics of US Supreme Court Edward Lee, Chase Broedersz, William Bialek We build simple models for the distribution of voting patterns in a group, using the Supreme Court of the United States as an example. The least structured, or maximum entropy, model that is consistent with the observed pairwise correlations among justices' votes is equivalent to an Ising spin glass. While all correlations (perhaps surprisingly) are positive, the effective pairwise interactions in the spin glass model have both signs, recovering some of our intuition that justices on opposite sides of the ideological spectrum should have a negative influence on one another. Despite the competing interactions, a strong tendency toward unanimity emerges from the model, and this agrees quantitatively with the data. The model shows that voting patterns are organized in a relatively simple ``energy landscape,'' correctly predicts the extent to which each justice is correlated with the majority, and gives us a measure of the influence that justices exert on one another. These results suggest that simple models, grounded in statistical physics, can capture essential features of collective decision making quantitatively, even in a complex political context. [Preview Abstract] |
Monday, March 3, 2014 11:27AM - 11:39AM |
B16.00002: Finding Structure in the ArXiv Alexander Alemi, Ricky Chachra, Paul Ginsparg, James Sethna We applied machine learning techniques to the full text of the arXiv articles and report a meaningful low-dimensional representation of this big dataset. Using Google's open source implementation of the continuous skip-gram model, word2vec, the vocabulary used in scientific articles is mapped to a Euclidean vector space that preserves semantic and syntactic relationships between words. This allowed us to develop techniques for automatically characterizing articles, finding similar articles and authors, and segmenting articles into their relevant sections, among other applications. [Preview Abstract] |
Monday, March 3, 2014 11:39AM - 11:51AM |
B16.00003: A scientific impact indicator based on the latent ``citability'' of a researcher's publications Joao Moreira, Xiaohan Zeng, Luis Amaral How to quantify the impact of a scientist's body of work is currently a matter of great concern. The use of bibliometric indicators, such as the \emph{h}-index or the Journal Impact Factor, have become widespread despite their known limitations. We surmise that many of the deficiencies of existing bibliometric indicators arise from their heuristic nature. Here, we pursue a principled approach to the development of an indicator to quantify the scientific impact of individual researchers, grounded on the functional form of the distribution of the ultimate number of citations. We validate our approach using the publication records of 1,283 researchers from seven scientific disciplines. Our approach has three distinct advantages. First, it accurately captures the overall scientific impact of researchers, as measured by ultimate citation counts. Second, in contrast to prior bibliometric indicators, our proposed measure does not depend on the number of publications, offering the possibility to compare researchers at different career stages. Third, more than other measures, our index is resistant to manipulation and rewards publication quality over quantity. [Preview Abstract] |
Monday, March 3, 2014 11:51AM - 12:03PM |
B16.00004: Stochastic Dynamics of Lexicon Learning in an Uncertain and Nonuniform World Richard Blythe, Rainer Reisenauer, Kenny Smith We study the time taken by a language learner to correctly identify the meaning of all words in a lexicon under conditions where many plausible meanings can be inferred whenever a word is uttered. We show that the most basic form of cross-situational learning - whereby information from multiple episodes is combined to eliminate incorrect meanings - can perform badly when words are learned independently and meanings are drawn from a nonuniform distribution. If learners further assume that no two words share a common meaning, we find a phase transition between a maximally efficient learning regime, where the learning time is reduced to the shortest it can possibly be, and a partially efficient regime where incorrect candidate meanings for words persist at late times. We obtain exact results for the word-learning process through an equivalence to a statistical mechanical problem of enumerating loops in the space of word-meaning mappings. [Preview Abstract] |
Monday, March 3, 2014 12:03PM - 12:15PM |
B16.00005: Statistical mechanics of human resource allocation Jun-ichi Inoue, He Chen We provide a mathematical platform to investigate the network topology of agents, say, university graduates who are looking for their positions in labor markets. The basic model is described by the so-called Potts spin glass which is well-known in the research field of statistical physics. In the model, each Potts spin (a tiny magnet in atomic scale length) represents the action of each student, and it takes a discrete variable corresponding to the company he/she applies for. We construct the energy to include three distinct effects on the students' behavior, namely, collective effect, market history and international ranking of companies. In this model system, the correlations (the adjacent matrix) between students are taken into account through the pairwise spin-spin interactions. We carry out computer simulations to examine the efficiency of the model. We also show that some chiral representation of the Potts spin enables us to obtain some analytical insights into our labor markets. [Preview Abstract] |
Monday, March 3, 2014 12:15PM - 12:27PM |
B16.00006: Gender differences in collaboration patterns Xiaohan Zeng, Jordi Duch, Marta Sales-Pardo, Filippo Radicchi, Haroldo V. Ribeiro, Teresa K. Woodruff, Luis A.N. Amaral Collaboration plays an increasingly important role in research productivity and impact. However, it remains unclear whether female and male researchers in science, technology, engineering and mathematical (STEM) disciplines differ significantly from each other in their collaboration propensity. Here, we report on an empirical analysis of the complete publication records of 3,920 faculty members in six STEM disciplines at selected top U.S. research universities. We find that while female faculty have significantly fewer co-authors over their careers, this can be fully explained by their lower number of publications. Indeed, we also find that females tend to distribute their co-authoring opportunities among their co-authors more evenly than males do. Our results suggest that females have had a greater propensity to collaborate, in order to succeed in a historically men-dominated academic world. Surprisingly, we find evidence that in molecular biology there has been a gender segregation within sub-disciplines. Female faculty in molecular biology departments tend to collaborate with smaller teams and publish in journals and fields where typical team size is smaller. Our results identify gender-specific collaborative behaviors as well as disciplines with distinct patterns. [Preview Abstract] |
Monday, March 3, 2014 12:27PM - 12:39PM |
B16.00007: The Regional Structure of Technical Innovation Dion O'Neale There is strong evidence that the productivity \emph{per capita} of cities and regions increases with population. One likely explanation for this phenomenon is that densely populated regions bring together otherwise unlikely combinations of individuals and organisations with diverse, specialised capabilities, leading to increased innovation and productivity. We have used the REGPAT patent database to construct a bipartite network of geographic regions and the patent classes for which those regions display a revealed comparative advantage. By analysing this network, we can infer relationships between different types of patent classes - and hence the structure of (patentable) technology. The network also provides a novel perspective for studying the combinations of technical capabilities in different geographic regions. We investigate measures such as the diversity and ubiquity of innovations within regions and find that diversity (resp. ubiquity) is positively (resp. negatively) correlated with population. We also find evidence of a nested structure for technical innovation. That is, specialised innovations tend to occur only when other more general innovations are already present. [Preview Abstract] |
Monday, March 3, 2014 12:39PM - 12:51PM |
B16.00008: A criterion for condensation in kinetically constrained one-dimensional transport models Daniel Miedema In transport, increasing the number of transporting particles not necessarily results in an increase of the throughput. When the density of a complex system increases, the current can decrease rapidly due to jamming effects. Jammed particles can form many clusters or one big cluster: a condensate in real space. We study condensation in one-dimensional transport models with a kinetic constraint. We find the conditions under which the arrested clusters can grow to a macroscopic condensate of arrested particles. We apply our finding to the well-known Nagel-Schreckenberg traffic flow model to analytically proof the existence of a condensate in a deterministic limit of this model, and verify this result with simulations. These results provide insight into dynamic arrest and dynamic phase separation in one-dimensional traffic and transport. [Preview Abstract] |
Monday, March 3, 2014 12:51PM - 1:03PM |
B16.00009: Competing effects of social balance and influence P. Singh, S. Sreenivasan, B. Szymanski, G. Korniss The theory of social balance is one of the key drivers of social dynamics. We study a model of social interactions in which the dynamics of social balance is competing with external influence. In this model, each node in a social network is in one of the three possible states - leftist, rightist, centrist. Only a link between two unequal extremist nodes is considered unfriendly and a triangle is balanced if it contains even number of unfriendly links. Thus triangles formed by a centrist, a leftist and a rightist are unbalanced. In this model, at each time step with probability $p$, we pick a random node and convert it into a centrist while with probability $(1-p)$, a randomly picked triangle is checked for balance and if needed, it is balanced by updating the state of one of the nodes in the triangle. We find that there exists a critical value $p_c$ such that for $p |
Monday, March 3, 2014 1:03PM - 1:15PM |
B16.00010: A network approach in analysis of the matching hypothesis Tao Jia, Robert Spivey, Gyorgy Korniss, Boleslaw Szymanski The matching hypothesis in social psychology claimed that people are more likely to form a committed relationship with someone who is equally attractive. This phenomenon can be well interpreted by the principle of homophily that people are apt to get in touch with others similar to them. Yet, social experiments indicate that people in general tend to prefer more attractive individuals regardless of their own attractiveness. Here study the stochastic matching process for different underlying networks and different attractiveness distributions. We showed that the correlation of attractiveness within couples could purely due to the limited number of acquaintance each person has and such correlation decreases as the network becomes more sparse. We also analyzed the effect of the degree distribution and the attractiveness on the number of individuals that can not find their partners. [Preview Abstract] |
Monday, March 3, 2014 1:15PM - 1:27PM |
B16.00011: Scale-free networks with temporary link deactivation for disease avoidance Leah Shaw, Maxim Shkarayev, Ilker Tunc We study epidemic spread on scale-free networks in which nodes can temporarily deactivate their links to infected neighbors and reactivate when their neighbors recover. We find that the topology of the subnetwork consisting of active links is fundamentally different from the original network topology, and we predict the scaling exponent of the active degree distribution. Further, we derive an improved low dimensional system of mean-field equations for dynamics of nodes and links based on the distribution of a node's neighbors conditioned on the total degree. [Preview Abstract] |
Monday, March 3, 2014 1:27PM - 1:39PM |
B16.00012: Internal and external influence in the US stock market Stanislav Borysov, Yasser Roudi, Alexander Balatsky We analyze the multivariate distribution of the US stock returns using pairwise interaction models, inspired by Ising models in glasses and neural networks. Using the inference methods from neural networks analysis we find unique descriptors of the dynamics of stock returns in periods of crisis. Our findings suggest that the near crash dynamics is primarily governed by external factors (external fields), while internal network structure (J couplings) are not significantly affected. [Preview Abstract] |
Monday, March 3, 2014 1:39PM - 1:51PM |
B16.00013: Can Quantum Physics Find the Answer to the World Financial Crisis. Lamine Dieng We assume the global wealth of nations within the G-Global to be an American call option described as a stochastic process. We let the American call option to grow and to eventually generate profits to the nations of the G-Global. We show profits taken to be a discontinuous process, because when an investment banker or a country makes more profits continuously, then their vision will be guided by greed. When banks try to maximize profits continuously and so they operate on the edge of bankruptcy. We also assume the global wealth to be an index defined in terms of the expected global wealth of nations and normalized by their GDPs. We impose the following conditions: a). The sum of the GDPs of all nations making the G-Global is one (1), the normalizing GDP should not have an influence on the global wealth. All nations should be treated on the same footing. b). the change of the global wealth of nations to be a supermartingale. We set the drift term of the expectation decreasing process to be equal to zero. We obtain an Ordinary Differential Equation describing the dynamic of global wealth [Preview Abstract] |
Monday, March 3, 2014 1:51PM - 2:03PM |
B16.00014: Prospect Theory for Online Financial Trading Yang-Yu Liu, Jose C. Nacher, Tomoshiro Ochiai, Mauro Martino, Yaniv Altshuler Prospect theory is widely viewed as the best available descriptive model of how people evaluate risk in experimental settings. According to prospect theory, people make decisions based on the potential value of losses and gains rather than the final outcome. People are risk-averse with respect to gains and risk-seeking with respect to losses, a phenomenon called ``loss aversion''. Despite of the fact that prospect theory has been well studied in behavioral economics at the theoretical level, there exist very few empirical research and most of them has been undertaken with micro-panel data. Here we analyze the trading activities of over 1.5 million members of an online financial trading community over 28 months, aiming to explore the large-scale empirical aspect of prospect theory. By analyzing and comparing the behaviour of ``winners'' and ``losers'', i.e., traders with positive or negative final net profit, we find clear evidence of the loss aversion phenomenon, an essence in prospect theory. This work demonstrates an unprecedented large-scale empirical evidence of prospect theory. It has immediate implication in financial trading, e.g., developing new trading strategies by minimizing the effect of loss aversion. It also provides opportunity to augment online social trading, where users are allowed to watch and follow the trading activity of others, by predicting potential winners based on their historical trading behaviour. [Preview Abstract] |
Monday, March 3, 2014 2:03PM - 2:15PM |
B16.00015: Finding hidden periodic signals in time series -- an application to stock prices Michael O'Shea Data in the form of time series appear in many areas of science. In cases where the periodicity is apparent and the only other contribution to the time series is stochastic in origin, the data can be `folded' to improve signal to noise and this has been done for light curves of variable stars with the folding resulting in a cleaner light curve signal. Stock index prices versus time are classic examples of time series. Repeating patterns have been claimed by many workers and include unusually large returns on small-cap stocks during the month of January, and small returns on the Dow Jones Industrial average (DJIA) in the months June through September compared to the rest of the year. Such observations imply that these prices have a periodic component. We investigate this for the DJIA. If such a component exists it is hidden in a large non-periodic variation and a large stochastic variation. We show how to extract this periodic component and for the first time reveal its yearly (averaged) shape. This periodic component leads directly to the `Sell in May and buy at Halloween' adage. We also drill down and show that this yearly variation emerges from approximately half of the underlying stocks making up the DJIA index. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700