arXiv subjects

Kan Chen

Publications and source records attributed to Kan Chen.

99 records · Page 6Linked to original sources

arXiv2004

Theory of the Three-Group Evolutionary Minority Game

Based on the adiabatic theory for the evolutionary minority game (EMG) that we proposed earlier[1], we perform a detail analysis of the EMG limited to three groups of agents. We derive a formula for the critical point of the transition from segregation (into opposing groups) to clustering (towards cautious behaviors). Particular to the three-group EMG, the strategy switching in the "extreme" group does not occur at every losing step and is strongly intermittent. This leads to an correction to the critical value of the number of agents at the transition, $N_c$. Our expression for $N_c$ is in agreement with the results obtained from our numerical simulations.

cond-mat.stat-mech↗

arXiv2003

Theory of Phase Transition in the Evolutionary Minority Game

We discover the mechanism for the transition from self-segregation (into opposing groups) to clustering (towards cautious behaviors) in the evolutionary minority game (EMG). The mechanism is illustrated with a statistical mechanics analysis of a simplified EMG involving three groups of agents: two groups of opposing agents and one group of cautious agents. Two key factors affect the population distribution of the agents. One is the market impact (the self-interaction), which has been identified previously. The other is the market inefficiency due to the short-time imbalance in the number of agents using opposite strategies. Large market impact favors "extreme" players who choose fixed strategies, while large market inefficiency favors cautious players. The phase transition depends on the number of agents ($N$), the reward-to-fine ratio ($R$), as well as the wealth reduction threshold ($d$) for switching strategy. When the rate for switching strategy is large, there is strong clustering of cautious agents. On the other hand, when $N$ is small, the market impact becomes large, and the extreme behavior is favored.

cond-mat↗

arXiv2000

Solitons in the one-dimensional forest fire model

Fires in the one-dimensional Bak-Chen-Tang forest fire model propagate as solitons, resembling shocks in Burgers turbulence. The branching of solitons, creating new fires, is balanced by the pair-wise annihilation of oppositely moving solitons. Two distinct, diverging length scales appear in the limit where the growth rate of trees, $p$, vanishes. The width of the solitons, $w$, diverges as a power law, $1/p$, while the average distance between solitons diverges much faster as $ d \sim \exp(π^2/12p)$.

cond-mat.stat-mech↗

arXiv2000

Scale Dependent Dimension in the Forest Fire Model

The forest fire model is a reaction-diffusion model where energy, in the form of trees, is injected uniformly, and burned (dissipated) locally. We show that the spatial distribution of fires forms a novel geometric structure where the fractal dimension varies continuously with the length scale. In the three dimensional model, the dimensions varies from zero to three, proportional with $log(l)$, as the length scale increases from $l \sim 1$ to a correlation length $l=ξ$. Beyond the correlation length, which diverges with the growth rate $p$ as $ξ \propto p^{-2/3}$, the distribution becomes homogeneous. We suggest that this picture applies to the ``intermediate range'' of turbulence where it provides a natural interpretation of the extended scaling that has been observed at small length scales. Unexpectedly, it might also be applicable to the spatial distribution of luminous matter in the universe. In the two-dimensional version, the dimension increases to D=1 at a length scale $l \sim 1/p$, where there is a cross-over to homogeneity, i. e. a jump from D=1 to D=2.

cond-mat.stat-mech↗

arXiv2000

Scale Dependent Dimension of Luminous Matter in the Universe

We present a geometrical model of the distribution of luminous matter in the universe, derived from a very simple reaction-diffusion model of turbulent phenomena. The apparent dimension of luminous matter, $D(l)$, depends linearly on the logarithm of the scale $l$ under which the universe is viewed: $D(l) \sim 3\log(l/l_0)/\log(ξ/l_0)$, where $ξ$ is a correlation length. Comparison with data from the SARS red-shift catalogue, and the LEDA database provides a good fit with a correlation length $ξ\sim 300$ Mpc. The geometrical interpretation is clear: At small distances, the universe is zero-dimensional and point-like. At distances of the order of 1 Mpc the dimension is unity, indicating a filamentary, string-like structure; when viewed at larger scales it gradually becomes 2-dimensional wall-like, and finally, at and beyond the correlation length, it becomes uniform.

astro-ph↗

arXiv1998

Dynamics of Dry Friction: A Numerical Investigation

We perform extended numerical simulation of the dynamics of dry friction, based on a model derived from the phenomenological description proposed by T. Baumberger et al.. In the case of small deviation from the steady sliding motion, the model is shown to be equivalent to the state- and rate-dependent friction law which was first introduced by Rice and Ruina on the basis of experiments on rocks. We obtain the dynamical phase diagram that agrees well with the experimental results on the paper-on-paper systems. In particular, the bifurcation between stick-slip and steady sliding are shown to change from a direct (supercritical) Hopf type to an inverted (subcritical) one as the driving velocity increases, in agreement with the experiments.

cond-mat↗

arXiv1997

Kondo Effect in Fermi Systems with a Gap: A Renormalization Group Study

We present the results of a Wilson Renormalization Group study of the single-impurity Kondo and Anderson models in a system with a gap in the conduction electron spectrum. The behavior of the impurity susceptibility and the zero-frequency response function, $T< >$ are discussed in the cases with and without particle-hole symmetry. In addition, for the asymmetric Anderson model the correlation functions, $<\vec S \cdot\vec σ(0)>$,$ $, and $ $ are computed.

cond-mat.str-el↗

arXiv1997

A general learning algorithm for solving optimization problems and its application to the spin glass problem

We propose a general learning algorithm for solving optimization problems, based on a simple strategy of trial and adaptation. The algorithm maintains a probability distribution of possible solutions (configurations), which is updated continuously in the learning process. As the probability distribution evolves, better and better solutions are shown to emerge. The performance of the algorithm is illustrated by the application to the problem of finding the ground state of the Ising spin glass. A simple theoretical understanding of the algorithm is also presented.

adap-org↗

arXiv1996

A Simple Learning Algorithm for the Traveling Salesman Problem

We propose a learning algorithm for solving the traveling salesman problem based on a simple strategy of trial and adaptation: i) A tour is selected by choosing cities probabilistically according to the ``synaptic'' strengths between cities. ii) The ``synaptic'' strengths of the links that form the tour are then enhanced (reduced) if the tour length is shorter (longer) than the average result of the previous trials. We perform extensive simulations of the random distance traveling-salesman problem. For sufficiently slow learning rates, near optimal tours can be obtained with the average optimal tour lengths close to the lower bounds for the shortest tour lengths.

adap-org↗