Searcharxiv⌕ Search

arXiv subjects

P. M. Hui

Publications and source records attributed to P. M. Hui.

47 records · Page 3Linked to original sources
arXiv2000

Segregation in a competing and evolving population

We study a recently proposed model in which an odd number of agents are competing to be in the minority. The agents have one strategy in hand which is to follow the most recent history. Each agent is also assigned a value p, which is the probability that an agent will follow the trend. Evolution is introduced through the modification of the value of p when the performance of an agent becomes unsatisfactory. We present numerical results for the distribution of p values in the population as well as the average duration between modifications at a given p for different values of the parameters in the model. We also point out the difference between the present model and a slightly modified model in which a strategy is randomly assigned to every agent initially.

cond-mat↗
arXiv1999

Evolutionary freezing in a competitive population

We show that evolution in a population of adaptive agents, repeatedly competing for a limited resource, can come to an abrupt halt. This transition from evolutionary to non-evolutionary behavior arises as the global resource level is changed, and is reminiscent of a phase transition to a frozen state. Its origin lies in the inductive decision-making of the agents, the limited global information that they possess and the dynamical feedback inherent in the system.

cond-mat.dis-nn↗
arXiv1999

Minority game with arbitrary cutoffs

We study a model of a competing population of N adaptive agents, with similar capabilities, repeatedly deciding whether to attend a bar with an arbitrary cutoff L. Decisions are based upon past outcomes. The agents are only told whether the actual attendance is above or below L. For L-> N/2, the game reproduces the main features of Challet and Zhang's minority game. As L is lowered, however, the mean attendances in different runs tend to divide into two groups. The corresponding standard deviations for these two groups are very different. This grouping effect results from the dynamical feedback governing the game's time-evolution, and is not reproduced if the agents are fed a random history.

cond-mat↗
arXiv1999

Enhanced winnings in a mixed-ability population playing a minority game

We study a mixed population of adaptive agents with small and large memories, competing in a minority game. If the agents are sufficiently adaptive, we find that the average winnings per agent can exceed that obtainable in the corresponding pure populations. In contrast to the pure population, the average success rate of the large-memory agents can be greater than 50 percent. The present results are not reproduced if the agents are fed a random history, thereby demonstrating the importance of memory in this system.

cond-mat↗
arXiv1998

Parallel updating cellular automaton models of driven diffusive Frenkel-Kontorova-type systems

Three cellular automaton (CA) models of increasing complexity are introduced to model driven diffusive systems related to the generalized Frenkel-Kontorova (FK) models recently proposed by Braun [Phys.Rev.E58, 1311 (1998)]. The models are defined in terms of parallel updating rules. Simulation results are presented for these models. The features are qualitatively similar to those models defined previously in terms of sequentially updating rules. Essential features of the FK model such as phase transitions, jamming due to atoms in the immobile state, and hysteresis in the relationship between the fraction of atoms in the running state and the bias field are captured. Formulating in terms of parallel updating rules has the advantage that the models can be treated analytically by following the time evolution of the occupation on every site of the lattice. Results of this analytical approach are given for the two simpler models. The steady state properties are found by studying the stable fixed points of a closed set of dynamical equations obtained within the approximation of retaining spatial correlations only upto two nearest neighboring sites. Results are found to be in good agreement with numerical data.

cond-mat↗
arXiv1998

Analytical Results For The Steady State Of Traffic Flow Models With Stochastic Delay

Exact mean field equations are derived analytically to give the fundamental diagrams, i.e., the average speed - car density relations, for the Fukui-Ishibashi one-dimensional traffic flow cellular automaton model of high speed vehicles $(v_{max}=M>1) $ with stochastic delay. Starting with the basic equation describing the time evolution of the number of empty sites in front of each car, the concepts of inter-car spacings longer and shorter than $M$ are introduced. The probabilities of having long and short spacings on the road are calculated. For high car densities $(ρ\geq 1/M)$, it is shown that inter-car spacings longer than $M$ will be shortened as the traffic flow evolves in time, and any initial configurations approach a steady state in which all the inter-car spacings are of the short type. Similarly for low car densities $(ρ\leq 1/M)$, it can be shown that traffic flow approaches an asymptotic steady state in which all the inter-car spacings are longer than $M-2$. The average traffic speed is then obtained analytically as a function of car density in the asymptotic steady state. The fundamental diagram so obtained is in excellent agreement with simulation data.

cond-mat↗
arXiv1998

Volatility and Agent Adaptability in a Self-Organizing Market

We present results for the so-called `bar-attendance' model of market behavior: $p$ adaptive agents, each possessing $n$ prediction rules chosen randomly from a pool, attempt to attend a bar whose cut-off is $s$. The global attendance time-series has a mean near, but not equal to, $s$. The variance, or `volatility', can show a minimum with increasing adaptability of the individual agents.

cond-mat↗
arXiv1997

Bose-Einstein condensation of finite number of confined particles

The partition function and specific heat of a system consisting of a finite number of bosons confined in an external potential are calculated in canonical ensemble. Using the grand partition function as the generating function of the partition function, an iterative scheme is established for the calculation of the partition function of system with an arbitrary number of particles. The scheme is applied to finite number of bosons confined in isotropic and anisotropic parabolic traps and in rigid boxes. The specific heat as a function of temperature is studied in detail for different number of particles, different degrees of anisotropy, and different spatial dimensions. The cusp in the specific heat is taken as an indication of Bose-Einstein condensation (BEC).It is found that the results corresponding to a large number of particles are approached quite rapidly as the number of bosons in the system increases. For large number of particles, results obtained within our iterative scheme are consistent with those of the semiclassical theory of BEC in an external potential based on the grand canonical treatment.

cond-mat↗
arXiv1996

Cellular automata models of traffic flow along a highway containing a junction

We examine various realistic generalizations of the basic cellular automaton model describing traffic flow along a highway. In particular, we introduce a {\em slow-to-start} rule which simulates a possible delay before a car pulls away from being stationary. Having discussed the case of a bare highway, we then consider the presence of a junction. We study the effects of acceleration, disorderness, and slow-to-start behavior on the queue length at the entrance to the highway. Interestingly, the junction's efficiency is {\it improved} by introducing disorderness along the highway, and by imposing a speed limit.

cond-mat↗
arXiv1995

Upper Bounds for the Critical Car Densities in Traffic Flow Problems

In most models of traffic flow, the car density $p$ is the only free parameter in determining the average car velocity $\langle v \rangle$. The critical car density $p_c$, which is defined to be the car density separating the jamming phase (with $\langle v \rangle = 0$) and the moving phase (with $\langle v \rangle > 0$), is an important physical quantity to investigate. By means of simple statistical argument, we show that $p_c < 1$ for the Biham-Middleton-Levine model of traffic flow in two or higher spatial dimensions. In particular, we show that $p_{c} \leq 11/12$ in 2 dimension and $p_{c} \leq 1 - \left( \frac{D-1}{2D} \right)^D$ in $D$ ($D > 2$) dimensions.

adap-org↗