«Flow improvement caused by agents who ignore trafﬁc rules Seung Ki Baek 1, Petter Minnhagen 1, Sebastian Bernhardsson 1, Kweon Choi2, Beom Jun ...»
Flow improvement caused by agents who ignore trafﬁc rules
Seung Ki Baek 1, Petter Minnhagen 1, Sebastian Bernhardsson 1, Kweon Choi2, Beom Jun Kim 3
1. Department of Physics, Umea University, Umea, Sweden
2. Gyeonggi Science High School, Suwon, Korea
3. Department of Physics, Sungkyunkwan University, Suwon, Korea
A system of agents moving along a road in both directions is studied within a cellular-automata
formulation. An agent steps to the right with probability q or to the left with 1−q when encountering others. Our model is restricted to two agent types, trafﬁc-rule abiders (q = 1) and ignorers (q = 1/2). The trafﬁc ﬂow is obtained as a function of density and relative fraction of the agent types.
The risk for jamming at a ﬁxed density, if starting from a disordered situation, is smaller when all the agents abide by the rule than when all ignore it. Nevertheless, the minimum occurs when a small fraction of ignorers are present within a majority of abiders. Characteristic spatial patterns are obtained and discussed.
Performance of strategy evaluation schemes for different price patterns Yongjoo Baek 1, Sang Hoon Lee 1, Hawoong Jeong 1,2
1. Department of Physics, Korea Advanced Institute of Science and Technology, Daejeon, Korea
2. Institute for the BioCentury, Korea Advanced Institute of Science and Technology, Daejeon, Korea We observe the performances of three strategy evaluation schemes, which are the history-dependent wealth game, the trend-opposing minority game, and the trend-following majority game in a stock market where the price is exogenously determined. The price is either directly adopted from the real stock market indices or generated with the Markov chain of order ≤ 2. The wealth game, as it learns from the history, shows relatively good performance unless the market is highly unpredictable. The majority game and the minority game are suitable for a market where their expectations are fulﬁlled. These observations suggest under which market circumstances each evaluation scheme is appropriate for modeling the behavior of real market traders.
Stock market volatility: An approach based on the concept of entropy Sonia R. Bentes1 and Rui Menezes2 ¹ISCAL, Lisboa Portugal, www.iscal.ipl.pt, email@example.com ²ISCTE, Lisboa, Portugal, www.iscte.pt, firstname.lastname@example.org One of the major issues studied in finance that had always intrigued, both scholars and practitioners, and to which there were not yet discovered a unified theory, is the reason why prices move over time and their underlying volatility, which actually seems to affect markets as a whole. Since there are several well known traditional techniques in the literature to measure stock market volatility, a central point in this debate that constitutes the actual scope of this paper, is to put together this common approach in which we discuss popular techniques like the variance and/or standard deviation, and an innovative methodology based on Econophysics which applies concepts of physics to explain economic/financial phenomena. In this particular study, we use the concepts of Shannon entropy, Renyi entropy and Tsallis entropy to capture the nature of volatility. More precisely, what we want to know in our study is if entropy is able to detect volatility in stock market indexes and to compare its values with the ones obtained from the variance and/or standard deviation analysis. For our purpose, we shall basically focus on the behaviour of seven stock market indexes: TSX 60 (Canada), CAC 40 (France), DAX 30 (Germany), MIB 30 (Italy) NIKKEI 225 (Japan), FTSE 100 (UK) and S&P 500 (USA) for a comparative analysis between the approaches mentioned above. The results are however mixed.
Impact of interaction structures on priority-queue network dynamics Won-kuk Cho, Byungjoon Min, K.-I. Goh, and I.-M. Kim Department of Physics, Korea University, Seoul, Korea Human activity patterns have been recently shown to follow heavy-tailed distributions. The priority-based queueing system has been proposed as a framework to treat that, and the dynamics of binary interacting priority-queue model was studied. We extend the study of the waiting time dynamics for priority-queue networks considering various forms of human interaction. The waiting time distributions exhibit power-law behaviours, with different exponents depending on specific interaction rules. Especially, the pairwise interactions seem to be essential dynamic consequences of the interactions in the priority-queue network dynamics. We also find out that the reciprocity of influence is relevant factor for the priority-queue network dynamics.
Long-memory covariance and correlation matrices and their use in ﬁnance
1. Swissquote Bank, Gland (Geneva Ofﬁce), Switzerland ´´
2. Ecole Polytechnique Federale de Lausanne (EPFL), Lausanne, Switzerland
3. Abdus Salam International Center for Theoretical Physics, Trieste, Italy Weighted covariance and correlation matrices exploit the time-clustering of events to improve the quality of forecasters used for ﬁnancial risk assessment and portfolio allocation. The underlying idea is to put more weight on recent events, as the latter are better predictors of the near future than events that occurred far in the past. As recently shown by econophysicists, standard weighted estimators using exponentially decreasing proﬁles rapidly lead to ill-conditioned matrices as the dimensionality of the estimator (i.e. the number of assets) increases. In part to overcome this issue, we introduce a new class of weighted covariance and correlation matrices C with adjustable long-memory proﬁles. Extending a famous result of Random Matrix Theory by Marˇ enko c and Pastur, we show how to preserve the conditioning and spectral noise band of C, while at the same time taking advantage of the clustered dynamics of the volatility for better covariance forecasts. We apply our results to improve Markovitz’s well-known Mean-Variance Optimization scheme.
Order-parameter ﬂows in stochastic processes of quantum information processing Jun-ichi Inoue Hokkaido University, Sapporo, Japan In terms of the stochastic process of quantum-mechanical version of Markov chain Monte Carlo method, we analytically derive macroscopically deterministic ﬂow equations of order parameters such as spontaneous magnetization in inﬁnite-range (d(= ∞)-dimensional) quantum spin systems. By means of the Trotter decomposition, we consider the transition probability of glaubertype dynamics of microscopic states for the corresponding (d + 1)-dimensional classical system.
Under the static approximation, differential equations with respect to macroscopic order parameters are explicitly obtained from the Master equation that describes the microscopic-law. In the steady state, we show that the equations are identical to the saddle point equations for the equilibrium state of the same system. The equation for the dynamical Ising model is recovered in the classical limit. We also check the validity of the static approximation by making use of computer simulations for ﬁnite size systems and discuss several possible extensions of our approach to disordered spin systems for statistical-mechanical informatics. Especially, we shall use our procedure to evaluate the decoding process of Bayesian image restoration. With the assistance of the concept of dynamical replica theory, we derive the zero-temperature ﬂow equation of image restoration measure showing some ‘non-monotonic’ behaviour in its time evolution. Other possible applications of our procedure to information retrieval process in the presence of quantum ﬂuctuation such as memory recalling process of quantum-mechanical version of the Hopﬁeld model will be discussed.
Dynamics of pedestrians: crowds and individuals
Human crowds and pedestrian groups exhibit complex and coordinated spatio-temporal patterns such as the spontaneous spatial organization of pedestrian ﬂows into lines, and the oscillations of ﬂuxes at gates or intersections. Despite their importance, these phenomena are not well understood, in particular the ‘microscopic’ interactions between the individuals and with their environment which govern the macroscopic behavior at medium and high densities.
In the frame of a collective project implying four French laboratories (LPT in Orsay, CRCA and IMT in Toulouse, BUNRAKU in Rennes), we have started an experimental and theoretical study of the formation of spatio-temporal structures within moving pedestrians crowds. We shall present the ﬁrst results from the experimental campaign of 2009. Our aim is to better understand the role of the various (physical and behavioral) parameters which control and modulate these structures in controlled laboratory conditions, and to develop realistic analytical and simulation models of crowds based on these experimental data.
Consistent Community Identification in Large Scale Networks Daniel Kim, Jinyoung You, Haewoon Kwak, Sue Moon, and Hawoong Jeong
1. Department of Physics, KAIST, Daejeon, Korea
2. Computer Science department, KAIST, Daejeon, Korea Various algorithms based on diverse measures have been suggested for mining communities in networks. Only a few of them, however, can be practically used for large scale networks and nevertheless suffer from inconsistent outcomes. We observe that approximately 40% of nodes in two social networking sites, Orkut and Cyworld with heterogeneous community size distributions, are grouped into communities in inconsistent ways in contrast to the results of other data by using our iterative reinforcing method. To find out the cause of the inconsistency, we applied our method to all the possible connected graphs up to the number of 9 nodes. We also obtain "modularity landscape" of all the possible community partitions in a specific 8 nodes network to see how our reinforcing method quantitatively advances. Finally, we compare our method with a modularity-free hierarchical link clustering method and discuss its validity and limitation.
The effect of the underlying topology on the synchronization of discrete-event simulation Jung Hwa Kim,Soonhyung Yook, Yup Kim
1. Department of Physics, Kyung Hee University, Seoul, Korea
2. Department of Physics, Kyung Hee University, Seoul, Korea
3. Department of Physics, Kyung Hee University, Seoul, Korea The parallel discrete-event simulation scheme is known to be closely related to the interface roughening phenomena. In this study, we investigate the Sneppen model without quenched noise on small-world networks. The Sneppen model without quenched noise belongs to Kadar-Parisi-Zhang universality class. To investigate the effect of underlying topology on the roughening phenomena of the model, we use small-world networks generated by adding shortcuts between randomly selected sites in one-dimensional lattice. From Monte Carlo simulations, we find that the growth exponent, β, crossovers from 1/3 to 1 and the roughness exponent, α, approaches to 1, when the number of shortcuts are finite. By measuring the height-height correlation function, we show that the shortcuts do a role of defects which cause such a nontrivial behavior.
Exploring the temporal correlation structure of a singluar return Gyuchang Lim, Soo Yong Kim, Kyungsik Kim
1. Department of Physics, Korea Advanced Institute of Science and Technology, Daejeon 305-701, South Korea
2. Department of Physics, Pukyong National University, Busan 608-737, Korea In this work, we construct a cross-correlation matrix from a set of time series segments, which are obtained by applying the moving window method to a singular time series. We assume that each segment represents the state of a system because fluctuations are closely related to the state of a system. As the volatility clustering is a well-known stylized fact for a financial market, it is in fact described that a certain pattern of fluctuations can be repeated regularly or irregulary. From these facts and assumptions, we find the group of segments which are statistically close to each other in terms of correlation coefficients. Particularly, since we give an appropriate probability measure to each group, we simulate the future returns not from a statistical model or a complex random walk process, but from empirically obtained returns.
The Survival, Persistence, and Correlation in a Stock Market Doo Hwan Kim, Moon-Yong Cha, and Jae Woo Lee
1. Department of Physics, Inha University, Incheon, Korea
We consider the survival, persistence, and the price-price correlation function in a Korean stock market. The survival probability, S (t ), measures the probability of a stock’s index remaining above (or below) a reference value up to a time interval t. We observed the survival probability followed a power law, S (t ) t, where the exponent depends on the reference level. The persisting time defined a time interval when the index remains above (or below) an initial index. The persistence probability also followed a power law, P(t ) t , with the persistent exponent 0.477 (2). We investigated the price-price correlation function, Fq (t ). The price-price correlation function followed a power law, Fq (t ) t h ( q ), where h (q ) is the generalized Hurst exponent. We observed a relation h(2) 1 within error bar.
Cascading failure model of world economy system Kyu-Min Lee 1, Jae-Suk Yang2, Gunn Kim3, Jaesung Lee4, Kwang-il Goh1, In-mook Kim1
1. Department of Physics, Korea University, Seoul, Korea
2. Columbia Business School, Columbia University, New York, NY, USA
3. Department of Physics, Kyung Hee University, Seoul, Korea
4. Department of Mathematics, Sogang University, Seoul, Korea In recent years, global economic crisis has become more important matter. In addition, this globalized crisis spreading shows emergent collective dynamics due to the interdependence between countries. From this point of view, here we study a cascading failure model on top of a network of countries based on trading relations. We examine the individual country’s role in the crisis spreading and how it is affected by trade network structure. We also discuss implications of our results for the prospect of the ongoing globalization.
Global dynamic routing for scale-free networks Xiang Ling 1, Mao-Bin Hu 1, Rui Jiang 1, Yong-Hong Wu 2 and Qing-Song Wu 1