In-mook Kim

Weighted Scale-Free Network in Financial Correlations

While many scale-free (SF) networks have been introduced recently for complex systems, most of them are binary random graphs. Here we introduce a weighted SF network in associated with the cross-correlations in stock price changes among the S&P 500 companies, where all vertices (companies) are fully connected and each edge has nonuniform weight given by the covariance between the two returns connected, normalized by their volatilities. Influence-strength (IS) is defined as the sum of the weights on the edges incident upon a given vertex. Then the IS distribution in its absolute magnitude | q | exhibits a SF behavior, P I (| q |)∼| q | -η with the exponent η≈1.8(1).

The market efficiency in the stock markets

Abstract.We study the temporal evolutions of three stock markets; Standard and Poors 500 index, Nikkei 225 Stock Average, and the Korea Composite Stock Price Index. We observe that the probability density function of the log-return has a fat tail but the tail index has been increasing continuously in recent years. We have also found that the variance of the autocorrelation function, the scaling exponent of the standard deviation, and the statistical complexity decrease, but that the entropy density increases as time goes over time. We introduce a modified microscopic spin model and simulate the model to confirm such increasing and decreasing tendencies in statistical quantities. These findings indicate that these three stock markets are becoming more efficient.

Binary spreading process with parity conservation.

Recently there has been a debate concerning the universal properties of the phase transition in the pair contact process with diffusion (PCPD) 2A-->3A, 2A-->0. Although some of the critical exponents seem to coincide with those of the so-called parity-conserving universality class, it was suggested that the PCPD might represent an independent class of phase transitions. This point of view is motivated by the argument that the PCPD does not conserve parity of the particle number. In the present work we question what happens if the parity conservation law is restored. To this end, we consider the reaction-diffusion process 2A-->4A, 2A-->0. Surprisingly, this process displays the same type of critical behavior, leading to the conclusion that the most important characteristics of the PCPD is the use of binary reactions for spreading, regardless of whether parity is conserved or not.

Effect of long-range interactions in the conserved Kardar-Parisi-Zhang equation

The conserved Kardar-Parisi-Zhang equation in the presence of long-range nonlinear interactions is studied by the dynamic renormalization group method. The long-range effect produces new fixed points with continuously varying exponents and gives distinct phase transitions, depending on both the long-range interaction strength and the substrate dimension

Phase transition in a triplet process.

d

Lower and upper bounds for the anomalous diffusion exponent on Sierpinski carpets

. The long-range interaction makes the surface width less rough than that of the short-range interaction. In particular, the surface becomes a smooth one with a negative roughness exponent at the physical dimension d=2.

Agent-Based Approach for Revitalization Strategy of Knowledge Ecosystem

We argue that the reaction-diffusion process 3A-->4A,3A-->2A exhibits a different type of continuous phase transition from an active into an absorbing phase. Because of the upper critical dimension d(c)> or =4/3 we expect the phase transition in 1+1 dimensions to be characterized by nontrivial fluctuation effects.

Exact solutions of a restricted ballistic deposition model on a one-dimensional staircase.

The anomalous diffusion exponent dw of random walks on a family of Sierpinski carpets are studied by analytical and numerical methods. We construct an effective bulk resistor and then establish the lower and upper bounds for dw where the lower bound turns out to be the same as that obtained with bond-moving renormalization. Numerical simulations on a family of Sierpinski carpets confirm our bounds, and show strong dependence of dw on the lacunarity of the carpet.

Optimization of consensus time by combining the voter and the majority voter models on scale-free networks

We conceptualize knowledge as an intellectual infrastructure that helps to maximize efficiency from the viewpoint of ecosystems. The knowledge ecosystem includes people and organizations that participate in the production, distribution, and consumption of this knowledge and information, as well as interactions between participants. We built the agent-based computational model of the ecosystem and induced seven key revitalization conditions from the perspective of complexity science and ecosystem management. We analyzed the effects of these conditions on the knowledge ecosystem based by the simulation of the agent-based model. Our results suggest that the proper implementation of seven revitalization conditions, focusing on the recovery of the positive feedback loop in the knowledge ecosystem, is crucial for sustainable development.

Dynamics of an interface driven through random media: The effect of a spatially correlated noise

The surface structure of a restricted ballistic deposition model is examined on a one-dimensional staircase with free boundary conditions. In this model, particles can be deposited only at the steps of the staircase. We set up recurrence relations for the surface fluctuation width [ital W] using the generating function method. Steady-state solutions are obtained exactly given system size [ital L]. In the infinite-size limit, [ital W] diverges as [ital L][sup [alpha]] with the scaling exponent [alpha]=1/2. The dynamic exponent [beta]([ital W][similar to][ital t][sup [beta]]) is also found to be 1/2 by solving the recurrence relations numerically. This model can be viewed as a simple variant of the model which belongs to the Kardar-Parisi-Zhang (KPZ) universality class ([alpha][sub [ital K][ital P][ital Z]]=1/2, [beta][sub [ital K][ital P][ital Z]]=1/3). Comparing its deposition time scale with that of the single-step model, we argue that [beta] must be the same as [beta][sub [ital K][ital P][ital Z]]/(1[minus][beta][sub [ital K][ital P][ital Z]]), which is consistent with our finding.