Alcohol consumption among college students: An agent‐based computational simulation

An agent‐based C++ program was developed to model student drinking. Student‐agents interact with each other and are randomly subjected to good or bad drinking experiences, to stories of other students' experiences, and to peer pressure. The program outputs drinking rates as functions of time based on various input parameters. The intent of this project is to simulate alcohol use, eventually adding other drugs, and possibly creating a simulation game for use as an educational tool. © 2008 Wiley Periodicals, Inc. Complexity, 2009     

The effects of mental model formation on group decision making: An agent‐based simulation

We investigated dynamics of group decision making on complex problems when agents can form mental models of others from discussion history. Results indicated that as the agents' memory capacity increases, the group reaches superficial consensus more easily. Surprisingly, however, the shared mental model of the problem develops only within a limited area of the problem space, because incorporating knowledge from others into one's own knowledge quickly creates local agreement on where relevant solutions are, leaving other potentially useful solutions beyond the scope of discussion. The mechanisms stifling group‐level exploration and their implications for decision making research are discussed. © 2010 Wiley Periodicals, Inc. Complexity 16: 49–57, 2011     

Parameter estimation of an agent‐based stock price model

The influence of the behavior and strategies of traders on stock price formation has attracted much interest. It is assumed that there is a positive correlation between the total net demand and the price change. A buy order is expected to increase the price, whereas a sell order is assumed to decrease it. We perform data analysis based on a recently proposed stochastic model for stock prices. The model involves long‐range dependence, self‐similarity, and no arbitrage principle, as observed in real data. The arrival times of orders, their quantity, and their duration are created by a Poisson random measure. The aggregation of the effect of all orders based on these parameters yields the log‐price process. By scaling the parameters, a fractional Brownian motion or a stable Levy process can be obtained in the limit. In this paper, our aim is twofold; first, to devise statistical methodology to estimate the model parameters with an application on high‐frequency price data, and second, to validate the model by simulations with the estimated parameters. We find that the statistical properties of agent level behavior are reflected on the stock price, and can affect the entire process. Moreover, the price model is suitable for prediction through simulations when the parameters are estimated from real data. The methods developed in the present paper can be applied to frequently traded stocks in general. Copyright © 2013 John Wiley & Sons, Ltd.     

Allostatic load as a complex clinical construct: A case‐based computational modeling approach

Allostatic load (AL) is a complex clinical construct, providing a unique window into the cumulative impact of stress. However, due to its inherent complexity, AL presents two major measurement challenges to conventional statistical modeling (the field's dominant methodology): it is comprised of a complex causal network of bioallostatic systems, represented by an even larger set of dynamic biomarkers; and, it is situated within a web of antecedent socioecological systems, linking AL to differences in health outcomes and disparities. To address these challenges, we employed case‐based computational modeling (CBM), which allowed us to make four advances: (1) we developed a multisystem, 7‐factor (20 biomarker) model of AL's network of allostatic systems; (2) used it to create a catalog of nine different clinical AL profiles (causal pathways); (3) linked each clinical profile to a typology of 23 health outcomes; and (4) explored our results (post hoc) as a function of gender, a key socioecological factor. In terms of highlights, (a) the Healthy clinical profile had few health risks; (b) the pro‐inflammatory profile linked to high blood pressure and diabetes; (c) Low Stress Hormones linked to heart disease, TIA/Stroke, diabetes, and circulation problems; and (d) high stress hormones linked to heart disease and high blood pressure. Post hoc analyses also found that males were overrepresented on the High Blood Pressure (61.2%), Metabolic Syndrome (63.2%), High Stress Hormones (66.4%), and High Blood Sugar (57.1%); while females were overrepresented on the Healthy (81.9%), Low Stress Hormones (66.3%), and Low Stress Antagonists (stress buffers) (95.4%) profiles. © 2015 Wiley Periodicals, Inc. Complexity 21: 291–306, 2016     

Agent‐based simulation of an N‐person game with parabolic payoff functions

We report computer simulation experiments based on our agent‐based simulation tool to model a new N‐person game based on John Conway's Game of Life. The individual agents may choose between two behavior options: cooperation or defection. The payoff (reward/penalty) functions are given as two parabolas: one for each option. After a certain number of iterations, the behavior of the agents stabilizes to either a constant value or oscillates around such a value. The simulation's goal is to investigate the effects of intermediate behavior on a society of agents. We have performed a systematic investigation of this game for all six possible cases of the mutual positions of parabolic payoff functions crossing each other at two points: x = 0.3 and 0.7 where x is the ratio of the cooperation choice to the total number of agents in the agent's neighborhood. The global ratios X(t) of the total number of cooperators in the entire array of agents as functions of time (iterations) and the solutions of the game X_final as functions of X₀ were observed for each case for Pavlovian, greedy, and conformist agents. The solutions have predictable tendencies only when the neighborhood is the entire array of greedy or conformist agents. In all other cases unexpected properties emerge. © 2009 Wiley Periodicals, Inc. Complexity, 2010     

An intensity‐based approach for equity modeling

This paper analyzes an intensity‐based approach for equity modeling. We use the Cox–Ingersoll–Ross (CIR) process to describe the intensity of the firm's default process. The intensity is purposely linked to the assets of the firm and consequently is also used to explain the equity. We examine two different approaches to link assets and intensity and derive closed‐form expressions for the firms' equity under both models. We use the Kalman filter to estimate the parameters of the unobservable intensity process. We demonstrate our approach using historical equity time series data from Merrill Lynch. Copyright © 2011 John Wiley & Sons, Ltd.     

Evaluating an hepatic enzyme induction mechanism through coarse‐ and fine‐grained measurements of an in silico liver

Results from simulation experiments falsified the hypothesis that a uniform distribution of simulated drug passing through an in silico liver (ISL) will produce a uniform extent of enzyme induction (EI). Wet‐lab EI experiments, as formulated, are infeasible. The simulated EI is intended to have a hepatic counterpart. The ISL is synthetic, physiologically based, fine‐grained, and multi‐agent. It has been validated against in situ drug disposition data. We discuss methodological considerations regarding the phenomenal manifold, multi‐level observation, and manipulation of synthetic models and their referents. Interestingly, a lower probability of metabolism caused higher EI and, counter‐intuitively, more extraction. © 2008 Wiley Periodicals, Inc. Complexity, 2009     

Agent‐based model for economic impact of free software

This article describes the potential impact that free (i.e., open source) software can have on an existing commercial software market. A model for the software market is constructed in terms of autonomous agents, which represent the users, the companies, and the free software providers. The model specifies a reservation price for each user agent and develops a gradient learning strategy for revenue‐maximizing company agents. Simulations explore parameters such as the demand distribution, and the relative importance of market share, advertising and random effects in product visibility. Results from the case without free software show a prevalence of monopolies, which is consistent with other studies of high‐technology market economics. The effects of free software are not uniform, but are highly parameter dependent. A “capture region” is found in which free software eventually dominates the market. © 2003 Wiley Periodicals, Inc.     

A systematic analysis of the N‐person chicken game

We report computer simulation experiments based on our agent‐based simulation tool to model the multiperson Chicken dilemma game for the case when the agents are greedy simpletons who imitate the action of that of their neighbors who received the highest payoff for its previous action. The individual agents may cooperate with each other for the collective interest or may defect, i.e., pursue their selfish interests only. After a certain number of iterations the proportion of cooperators stabilizes to either a constant value or oscillates around such a value. The payoff (reward/penalty) functions are given as two straight lines: one for the cooperators and another for the defectors. The payoff curves are functions of the ratio of cooperators to the total number of agents. Even for linear payoff functions, we have four free parameters that determine the payoff functions that have the following properties: (1) Both payoff functions increase with the increasing number of cooperators. (2) In the region of low cooperation the cooperators have a higher reward than the defectors. (3) When the cooperation rate is high, there is a higher payoff for defecting behavior than for cooperating behavior. (4) As a consequence, the slope of the D function is greater than that of the C function and the two payoff functions intersect. (5) All agents receive a lower payoff if all defect than if all cooperate. We have investigated the behavior of the agents systematically. The results show that the solutions have predictable tendencies but they are nontrivial and quite irregular. The solutions show drastic changes in the parameter ranges 0.6 ≤ R ≤ 0.65 for all values of S and 0 ≤ S ≤ 0.2 when R < 0.6 (R is the reward for mutual cooperation and S is the sucker's payoff to a lonely cooperator). © 2010 Wiley Periodicals, Inc. Complexity, 2010     

Emergence and induction of cellular automata rules via probabilistic reinforcement paradigms

Two simple models of emergence and automated induction are described. In the first, an initially random process comes, over time, to emulate a deterministic process with noise. In the second, an induction algorithm is used to make unbiased best guess estimates of cellular automata rules generating a given time series of binary strings. The general conclusions are as follows: (1) that it may not be possible to distinguish between a stochastic process with selection and reinforcement and a noisy deterministic process; and (2) automated induction algorithms will often be vulnerable to errors of type 1 when faced with random data. In this second case, this leads to a method for study of the modeling class assumed in the induction algorithm. © 2006 Wiley Periodicals, Inc. Complexity 11: 44–57, 2006     

Hedging unit‐linked life insurance contracts in a financial market driven by shot‐noise processes

We consider the risk‐minimizing hedging problem for unit‐linked life insurance in a financial market driven by a shot‐noise process. Because the financial market is incomplete, the insurance claims cannot be hedged completely by trading stocks and bonds only, leaving some risk to the insurer. The theory of ((pseudo) locally) risk‐minimization is applied after a change of measure. Then the risk‐minimizing trading strategies and the associated intrinsic risk processes are determined for two types of unit‐linked contracts represented by the pure endowment and the term insurance. Copyright © 2009 John Wiley & Sons, Ltd.     

Nonintrusive reduced‐order modeling of parametrized time‐dependent partial differential equations

We propose a nonintrusive reduced‐order modeling method based on the notion of space‐time‐parameter proper orthogonal decomposition (POD) for approximating the solution of nonlinear parametrized time‐dependent partial differential equations. A two‐level POD method is introduced for constructing spatial and temporal basis functions with special properties such that the reduced‐order model satisfies the boundary and initial conditions by construction. A radial basis function approximation method is used to estimate the undetermined coefficients in the reduced‐order model without resorting to Galerkin projection. This nonintrusive approach enables the application of our approach to general problems with complicated nonlinearity terms. Numerical studies are presented for the parametrized Burgers' equation and a parametrized convection‐reaction‐diffusion problem. We demonstrate that our approach leads to reduced‐order models that accurately capture the behavior of the field variables as a function of the spatial coordinates, the parameter vector and time. © 2013 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq 2013     

Network theory and behavioral finance in a heterogeneous market environment

This article addresses the stock market as a complex system. The complexity of the stock market arises from the structure of the environment, agent heterogeneity, interactions among agents, and interactions with market regulators. We develop the idea of a meta‐model, which is a model of models represented in an agent‐based model that allows us to investigate this type of market complexity. The novelty of this article is the incorporation of various complexities captured by network theoretical models or induced by investment behavior. The model considers agents heterogeneous in terms of their strategies and investment behavior. Four investment strategies are included in the model: zero‐intelligence, fundamental strategy, momentum (trend followers), and adaptive trading strategy using the artificial neural network algorithm. In terms of behavior, the agents can be risk averse or loss occupied with overconfidence or conservative biases. The agents may interact with each other by sharing market sentiments through a structured scale‐free network. The market regulator controls the market through various control tools such as the risk‐free rate and taxation. Parameters are calibrated to the S&P500. The calibration is implemented using a scatter search heuristic approach. The model is validated using various stylized facts of stock return patterns such as excess kurtosis, auto‐correlation, and ARCH effect phenomena. Analysis at the macro and micro level of the market was performed by measuring the sensitivity of volatility and market capital and investigating the wealth distributions of the agents. We found that volatility is more sensitive to the model parameters than to market capital, and thus, the level of volatility does not affect market capital. In addition, the findings suggest that the efficient market hypothesis holds at the macro level but not at the micro level. © 2016 Wiley Periodicals, Inc. Complexity 21: 530–554, 2016     

Synchronization of memristor‐based delayed BAM neural networks with fractional‐order derivatives

This article deals with the problem of synchronization of fractional‐order memristor‐based BAM neural networks (FMBNNs) with time‐delay. We investigate the sufficient conditions for adaptive synchronization of FMBNNs with fractional‐order 0 < α < 1. The analysis is based on suitable Lyapunov functional, differential inclusions theory, and master‐slave synchronization setup. We extend the analysis to provide some useful criteria to ensure the finite‐time synchronization of FMBNNs with fractional‐order 1 < α < 2, using Mittag‐Leffler functions, Laplace transform, and linear feedback control techniques. Numerical simulations with two numerical examples are given to validate our theoretical results. Presence of time‐delay and fractional‐order in the model shows interesting dynamics. © 2016 Wiley Periodicals, Inc. Complexity 21: 412–426, 2016     

Co‐existence of Poisson and non‐Poisson processes in ordered parallel multilane pedestrian traffic

We show that non‐Poisson and Poisson processes can coexist in ordered parallel multilane pedestrian traffic, in the presence of lane switching which asymmetrically benefits the switchers and nonswitchers. Pedestrians join at the tail end of a queue and transact at the opposite front end. Their aim is to complete a transaction within the shortest possible time, and they can transfer to a shorter queue with probability p_s. Traffic is described by the utilization parameter U = λ〈t_s〉/N, where λ is the average rate of pedestrians entering the system, 〈t_s〉 is the average transaction time, and N is the number of lanes. Using an agent‐based model, we investigate the dependence of the average completion time 〈t_c〉 with variable K = 1 + (1 ? U)^?1 for different N and 〈t_s〉 values. In the absence of switching (p_s = 0), we found that 〈t_c〉 ∝ K^τ, where τ ≈ 1 regardless of N and 〈t_s〉. Lane switching (p_s = 1) reduces 〈t_c〉 for a given K, but its characteristic dependence with K differs for nonswitchers and switchers in the same traffic system. For the nonswitchers, 〈t_c〉 ∝ K^τ, where τ < 1. At low K values, switchers have a larger 〈t_c〉 that also increases more rapidly with K. At large K, the increase rates become equal for both. For nonswitchers, the possible t_c values obey an exponentially decaying probability density function p(t_c). The switchers on the other hand, are described by a fat‐tailed p(t_c) implying that a few are penalized with t_c values that are considerably longer than any of those experienced by nonswitchers. © 2006 Wiley Periodicals, Inc. Complexity 11: 35–42, 2006     

Cooperation in bike racing—When to work together and when to go it alone

We present an agent‐based model of bicycle racing that incorporates both physiology and the types of multiplayer scenarios that arise in real races. In these scenarios, riders can choose to share the workload with other riders (cooperate) or pursue their own self‐interests (defect). We compare the model's predictions to race situations and use it to investigate how different strategies can affect outcomes. We find that an individual player's best strategy depends on fitness level: below‐average riders fare better as defectors whereas above‐average riders perform better as cooperators. The strategies of stronger riders affect their teammates' results as well. The teammates of defecting strong riders fare worse overall than the teammates of cooperating strong riders. These results reproduce a dynamic that played out in the 2009 Tour de France. The winner, Alberto Contador, pursued a strategy perceived by many to be unusually uncooperative by repeatedly defecting on his teammates. The strategy worked to his advantage but may have negatively affected his teammates' placements. © 2011 Wiley Periodicals, Inc. Complexity, 17,39–44, 2011     

基于方向多尺度变换和统计建模的纹理分类方法

纹理是图像分析和识别中经常使用的关键特征, 而小波变换则是图像纹理表示和分类中的常用工具. 然而, 基于小波变换的纹理分类方法常常忽略了小波低频子带信息, 并且无法提取图像纹理的块状奇异信息. 本文提出小波子带系数的局部能量直方图建模方法、轮廓波特征的Poisson 混合模型建模方法和基于轮廓波子带系数聚类的特征提取方法, 并将其应用于图像纹理分类上. 基于局部能量直方图的纹理分类方法解决了小波低频子带的建模难题, 基于Poisson 混合模型的纹理分类方法则首次将Poisson 混合模型用于轮廓子带特征的建模, 而基于轮廓波域聚类的纹理分类方法是一种快速的分类方法. 实验结果显示, 本文所提出的三类方法都超过了当前典型的纹理分类方法.     

基于小波域的多尺度非参数图像分割方法

主要介绍了一种基于信息熵理论及图像多尺度信息来对图像进行非参数主动轮廓模型分割的有效方法.由于小波多分辨率特性的引入,可以最大程度地利用图像多尺度信息以确保分割的准确性和完整性.又由于小波变换的特性,低频信息的使用更是进一步降低了噪声影响.文中把图像分割问题定义为在分割区域边缘长度满足一定约束条件下,图像标记场与各个尺度图像像素值之间的互信息熵最大化过程.该方法可以有效地降低噪声对于分割的影响,及确保分割的准确性和完整性.     

Multiscale discontinuous Petrov–Galerkin method for the multiscale elliptic problems

In this article, we present a new multiscale discontinuous Petrov–Galerkin method (MsDPGM) for multiscale elliptic problems. This method utilizes the classical oversampling multiscale basis in the framework of a Petrov–Galerkin version of the discontinuous Galerkin method, allowing us to better cope with multiscale features in the solution. MsDPGM takes advantage of the multiscale Petrov–Galerkin method (MsPGM) and the discontinuous Galerkin method (DGM). It can eliminate the resonance error completely and decrease the computational costs of assembling the stiffness matrix, thus, allowing for more efficient solution algorithms. On the basis of a new H² norm error estimate between the multiscale solution and the homogenized solution with the first‐order corrector, we give a detailed convergence analysis of the MsDPGM under the assumption of periodic oscillating coefficients. We also investigate a multiscale discontinuous Galerkin method (MsDGM) whose bilinear form is the same as that of the DGM but the approximation space is constructed from the classical oversampling multiscale basis functions. This method has not been analyzed theoretically or numerically in the literature yet. Numerical experiments are carried out on the multiscale elliptic problems with periodic and randomly generated log‐normal coefficients. Their results demonstrate the efficiency of the proposed method.