A Simulation Framework for Heterogeneous Agents

We introduce a generic simulation framework suitable for agent-based simulations featuring the support of heterogeneous agents, hierarchical scheduling, and flexible specification of design parameters. One key aspect of this framework is the design specification: we use a format based on the Extendible Markup Language (XML) that is simple-structured yet still enables the design of flexible models. Another issue in agent-based simulations, especially when ready-made components are used, is the heterogeneity arising from both the agents' implementations and the underlying platforms. To tackle such obstacles, we introduce a wrapper technique for mapping the functionality of agents living in an interpreter-based environment to a standardized JAVA interface, thus facilitating the task for any control mechanism (like a simulation manager) because it has to handle only one set of commands for all agents involved. Again, this mapping is made by an XML-based definition format. We demonstrate the technique by applying it to a simple sample simulation of two mass marketing firms operating in an artificial consumer environment.     

A Role Based Model for the Normative Specification of Organized Collective Agency and Agents Interaction

In this article we propose a role based model for the specification of organized collective agency, based on the legal concept of artificial person and on the normative perspective of organizational systems. We focus on the analysis of groups of agents (humans or not) that want to act collectively in a (more or less) permanent basis, and in a stable and organized way, as it is the typical case of organizations. We argue that in those cases such groups of agents should give rise to a new agent, that we call of institutionalized agent, with its own identity, whose structure is essentially defined through the characterization of a set of roles and whose behavior is determined by the acts of the agents that play such roles. We also present a deontic and action modal logic that captures the concept of acting in a role and relates it with the deontic notions of obligation, permission and prohibition. This logic is used in the formal specification of institutionalized agents and of societies of agents and in the rigorous analysis of them. We pay particular attention to the interaction between agents through contracts or other normative relations. A high level specification language is also suggested.     

PRICE DYNAMICS, INFORMATIONAL EFFICIENCY, AND WEALTH DISTRIBUTION IN CONTINUOUS DOUBLE-AUCTION MARKETS

This paper studies the properties of the continuous double-auction trading mechanism using an artificial market populated by heterogeneous computational agents. In particular, we investigate how changes in the population of traders and in market microstructure characteristics affect price dynamics, information dissemination, and distribution of wealth across agents. In our computer-simulated market only a small fraction of the population observe the risky asset's fundamental value with noise, while the rest of the agents try to forecast the asset's price from past transaction data. In contrast to other artificial markets, we assume that the risky asset pays no dividend, thus agents cannot learn from past transaction prices and subsequent dividend payments. We find that private information can effectively disseminate in the market unless market regulation prevents informed investors from short selling or borrowing the asset, and these investors do not constitute a critical mass. In such case, not only are markets less efficient informationally, but may even experience crashes and bubbles. Finally, increased informational efficiency has a negative impact on informed agents' trading profits and a positive impact on artificial intelligent agents' profits.     

Modelling a Society of Simple Agents: From Conceptual Specification to Experimentation

Much research concerning the design of multi-agent systems (at a conceptual level) addresses complex agents that exhibit complex interaction patterns. Due to this complexity, it is difficult to perform rigorous experimentation. On the other hand, systematic experimental work regarding behaviour of societies of more simple agents, while reporting valuable results, often lacks conceptual specification of the system under consideration. In this paper, the compositional multi-agent modelling framework DESIRE is not only successfully used to develop a conceptual specification of the simple agents discussed by A. Cesta, M. Miceli and P. Rizzo (Lecture Notes in Artificial Intelligence, vol. 1038, Springer-Verlag: Berlin, pp. 128–138, 1996), but also to simulate the behaviour in a dynamical environment. In the DESIRE framework, a conceptual specification, which provides a high-level view of an agent, has enough detail for automatic prototype generation. The prototype implementation of the con-ceptual specification of the simple agents has been used to replicate, and extend, one of the experiments reported by Cesta et al.     

Specification and analysis of real-time problem solvers

The authors provide a method for the specification of real-time artificial intelligence (AI) problem solvers. Using this method, a formal specification of a real-time problem is presented. In addition, a method for analyzing real-time AI problem solvers is examined using a case study of two real-time problem solvers, namely DYNORAII and RTA* for the real-time path planning problem. New results on worst-case and average-case complexity of the problem, and of the algorithms that solve it, and an experimental evaluation of DYNORAII and RTA* for deadline compliance and response-time minimization are provided     

Behavior monitoring under uncertainty using Bayesian surprise and optimal action selection

The increasing trend towards delegating tasks to autonomous artificial agents in safety–critical socio-technical systems makes monitoring an action selection policy of paramount importance. Agent behavior monitoring may profit from a stochastic specification of an optimal policy under uncertainty. A probabilistic monitoring approach is proposed to assess if an agent behavior (or policy) respects its specification. The desired policy is modeled by a prior distribution for state transitions in an optimally-controlled stochastic process. Bayesian surprise is defined as the Kullback–Leibler divergence between the state transition distribution for the observed behavior and the distribution for optimal action selection. To provide a sensitive on-line estimation of Bayesian surprise with small samples twin Gaussian processes are used. Timely detection of a deviant behavior or anomaly in an artificial pancreas highlights the sensitivity of Bayesian surprise to a meaningful discrepancy regarding the stochastic optimal policy when there exist excessive glycemic variability, sensor errors, controller ill-tuning and infusion pump malfunctioning. To reject outliers and leave out redundant information, on-line sparsification of data streams is proposed.     

A Logical Model for Agent Communication Languages

The current ACL proposals show some shortcomings with respect to the definition of their semantics. Our paper aims at tackling those issues by defining an ACL semantics as a specification of the analytical effects of agent communicative acts. We analyze agent communication in terms of concepts taken from Speech Act Theory, as several researchers have already done, but move away from the mainstream view of artificial agent research, as we define communicative acts in terms of changes at the level of social relationship between agents. We take commitment to be a primitive concept underlying the social dimension of multiagent systems, and define a basic artificial institution whose aim is to provide agents with the means to affect the commitment network that binds them to each other.     

An Heterogeneous, Endogenous and Coevolutionary GP-Based Financial Market

Stock markets are very important in modern societies and their behavior has serious implications for a wide spectrum of the world's population. Investors, governing bodies, and society as a whole could benefit from better understanding of the behavior of stock markets. The traditional approach to analyzing such systems is the use of analytical models. However, the complexity of financial markets represents a big challenge to the analytical approach. Most analytical models make simplifying assumptions, such as perfect rationality and homogeneous investors, which threaten the validity of their results. This motivates alternative methods.In this paper, we report an artificial financial market and its use in studying the behavior of stock markets. This is an endogenous market, with which we model technical, fundamental, and noise traders. Nevertheless, our primary focus is on the technical traders, which are sophisticated genetic programming based agents that co- evolve (by learning based on their fitness function) by predicting investment opportunities in the market using technical analysis as the main tool. With this endogenous artificial market, we identify the conditions under which the statistical properties of price series in the artificial market resemble some of the properties of real financial markets. By performing a careful exploration of the most important aspects of our simulation model, we determine the way in which the factors of such a model affect the endogenously generated price. Additionally, we model the pressure to beat the market by a behavioral constraint imposed on the agents reflecting the Red Queen principle in evolution. We have demonstrated how evolutionary computation could play a key role in studying stock markets, mainly as a suitable model for economic learning on an agent- based simulation.     

Agent-Based Simulation and Microstructure Modeling of Immature Stock Markets

This work presents an artificial order-driven market able to reproduce mature and immature stock markets properties in the case of a single traded asset. This agent-based artificial market is designed to simulate characteristics of immature stock markets (high risk and low efficiency) by reproducing their stylized facts related mainly to information asymmetry and herd behavior. These two properties are modeled by combining social network and multi-agent simulations. The constructed scale-free social network, linking the modeled investors, gives rise to both informed and uninformed agents communities. Different assortative topologies are proposed and linked to different degrees of information asymmetry and market maturities. Several simulation experiments show that the modeled information asymmetry and herd behavior succeed in reproducing artificially some important stylized facts characterizing differences between immature and mature stock markets.     

Computational speed and high-frequency trading profitability: an ecological perspective

High-frequency traders (HFTs) account for a considerable component of equity trading but we know little about the source of their trading profits and to what extend IT based differentiators such as news processing power and ultra-low latency has contributed to competitive advantage within HFT realm. Given a fairly modest amount of empirical evidence on the subject, we study the effects of computational speed on HFTs’ profits through an experimental artificial agent-based equity market. Our approach relies on an ecological modelling inspired from the r/K selection theory, and is designed to assess the relative financial performance of two classes of aggressive HFT agents endowed with dissimilar computational capabilities. We use a discrete-event news simulation system to capture the information processing disparity and order transfer delay, and simulate the dynamics of the market. Through Monte Carlo simulation we obtain in our empirical setting robust estimates of the expected outcome.     

Automated traders in commodities markets: Case of producer–consumer institution

Automatizing commodities’ price negotiation was hard to achieve in practice, mainly because of logistical complications. The purpose of our work is to show that it is possible to automatize thoroughly commodities’ trading in the futures market by replacing human traders with artificial agents. As a starting step, we designed a market institution, called producer–consumer, where only an automated seller and an automated buyer can trade on behalf of the producer and consumer, respectively. The producer and consumer periodically feed their trading agents with supply and demand (S&D) forecasts. We suggested a parameterizable trading strategy, called bands and frequencies, for the agents. To measure the overall efficiency of this trading system in terms of price stability and liquidity, we made some hypotheses on the benchmark price curve and its linkages to S&D curves and other relevant market variables. Then we proposed analytical tools to measure strategy performance. Finally, we conducted some computer simulations to prove the workability of this approach.     

A MAS-based infrastructure for negotiation and its application to a water-right market

This paper presents a MAS-based infrastructure for the specification of a negotiation framework that handles multiple negotiation protocols in a coherent and flexible way. Although it may be used to implement one single type of agreement mechanism, it has been designed in such a way that multiple mechanisms may be available at any given time, to be activated and tailored on demand (on-line) by participating agents. The framework is also generic enough so that new protocols may be easily added. This infrastructure has been successfully used in a case study to implement a simulation tool as a component of a larger framework based on an electronic market of water rights.     

An empirical semantics approach to reasoning about communication

When designing multiagent systems, one can often avail of an existing specification of communication rules (in the form of protocols, ACL semantics, etc.). The question that then arises naturally is how to design appropriate agents to operate on such a specification. Moreover, if the multiagent system in question exhibits the characteristics of an open system, the problem is complicated even further by the fact that adherence to a supposedly agreed specification cannot be ensured on the side of other agents.This paper presents an architecture for dealing with a generic type of pre-specified communication patterns (containing surface structure and logical constraint specifications) based on an empirical semantics model of communication. By combining existing expectations about the use of communication with empirical observation, this model allows for a flexible adaptation to evolving communication semantics.The architecture itself is based on the InFFrA social reasoning framework and the concept of interaction frames. When interpreted according to the empirical semantics approach, interaction frames that represent classes of interaction situations can be used to conduct decision-theoretic reasoning about communication.After introducing the abstract architecture and giving a formal model for its probabilistic semantics, the results of an experimental validation of the approach in a complex domain are presented to illustrate its effectiveness.     

Relative risk aversion and wealth dynamics

As a follow-up to the work of Chen and Huang [S.-H. Chen, Y.-C. Huang, Risk preference, forecasting accuracy and survival dynamics: simulations based on a multi-asset agent-based artificial stock market, Working Paper Series 2004-1, AI-ECON Research Center, National Chengchi University, 2004; S.-H. Chen, Y.-C. Huang, Risk preference and survival dynamics, in: T. Terano, H. Kita, T. Kaneda, K. Arai, H. Deghchi (Eds.), Agent-Based Simulation: From Modeling Methodologies to Real-World Applications, Springer Series on Agent-Based Social Systems, vol. 1, 2005, pp. 135-143], this paper continues to explore the relationship between wealth share dynamics and risk preferences in the context of an agent-based multi-asset artificial stock market. We simulate a multi-asset agent-based artificial stock market composed of heterogeneous agents with different degrees of relative risk aversion. As before, we find that the difference in risk aversion and the resultant saving behavior are the primary forces in determining the survivability of agents. In addition to the stability of the saving behavior, the level of the saving rate also plays a crucial role. The agents with stable saving behavior, e.g., the log-utility agents, may still become extinct because of their low saving rates, whereas the agents with unstable saving behavior may survive because of their high saving rates, implied by their highly risk-averse preferences.     

基于GNP的多Agent人工股市模型

结合奥地利学派的经济思想，本文介绍了一种新的基于GNP算法的多Agent人工股市模型。该模型采用GNP算法来模拟交易个体的行为模式，进化他们的决策规则；同时在设计上强化Agent的异质性，并利用GA算法来优化模型参数。仿真结果表明，GNP-ASM模型表现出很好的统计性能，能够体现真实股市的一些基本特征。     

Anatomy of a forward-looking open standard [RapidIO]

The RapidIO interconnect specification represents the state of the art in standards. development. This software-transparent specification resolves an essential design bottleneck while also creating a clear migration path for systems based on existing or legacy architectures. The process of bringing this specification to market demonstrates that contributing to a collaborative standards development effort can be beneficial to all the participants. Designing to the specification can reduce manufacturing costs and time to market while significantly increasing bandwidth     

A representation theorem for minmax regret policies

Decision making under uncertainty is one of the central tasks of artificial agents. Due to their simplicity and ease of specification, qualitative decision tools are popular in artificial intelligence. Brafman and Tennenholtz [R.I. Brafman, M. Tennenholtz, An axiomatic treatment of three qualitative decision criteria, J. ACM 47 (3) (2000) 452-482] model an agent's uncertain knowledge as her local state, which consists of states of the world that she deems possible. A policy determines for each local state a total preorder of the set of actions, which represents the agent's preference over these actions. It is known that a policy is maximin representable if and only if it is closed under unions and satisfies a certain acyclicity condition.In this paper we show that the above conditions, although necessary, are insufficient for minmax regret and competitive ratio policies. A complete characterization of these policies is obtained by introducing the best-equally strictness.     

Platform design for large-scale artificial market simulation and preliminary evaluation on the K computer

Artificial market simulations have the potential to be a strong tool for studying rapid and large market fluctuations and designing financial regulations. High-frequency traders, that exchange multiple assets simultaneously within a millisecond, are said to be a cause of rapid and large market fluctuations. For such a large-scale problem, this paper proposes a software or computing platform for large-scale and high-frequency artificial market simulations (Plham: /pl\(\Lambda\)m). The computing platform, Plham, enables modeling financial markets composed of various brands of assets and a large number of agents trading on a short timescale. The design feature of Plham is the separation of artificial market models (simulation models) from their execution (execution models). This allows users to define their simulation models without parallel computing expertise and to choose one of the execution models they need. This computing platform provides a prototype execution model for parallel simulations, which exploits the variety in trading frequency among traders, that is, the fact that some traders do not require up-to-date information of markets changing in millisecond order. We evaluated a prototype implementation on the K computer using up to 256 computing nodes.     

Cooperative Tasking for Deterministic Specification Automata

This paper proposes necessary and sufficient conditions for task decomposability with respect to an arbitrary finite number of agents. It is furthermore shown that fulfilling the decomposed local tasks by individual agents guarantees the satisfaction of the original global decomposable task. A divide‐and‐conquer approach for cooperative tasking among multi‐agent systems is proposed. The basic idea is to decompose an assigned global specification (given as a deterministic automaton) into subtasks for individual concurrent agents such that the fulfillment of these subtasks by each individual agent leads to the satisfaction of the global specification as a team. A cooperative scenario involving three robots has been implemented to illustrate the proposed technique. This work provides insights into what kinds of tasks can be achieved distributively, which helps designers specify achievable global tasks for a group of agents and design necessary information sharing among each other for a particular task.