Robust Equilibria in Indefinite Linear-Quadratic Differential Games

Equilibria in dynamic games are formulated often under the assumption that the players have full knowledge of the dynamics to which they are subject. Here, we formulate equilibria in which players are looking for robustness and take model uncertainty explicitly into account in their decisions. Specifically, we consider feedback Nash equilibria in indefinite linear-quadratic differential games on an infinite time horizon. Model uncertainty is represented by a malevolent input which is subject to a cost penalty or to a direct bound. We derive conditions for the existence of robust equilibria in terms of solutions of sets of algebraic Riccati equations.     

Zero-Sum Stochastic Games with Partial Information

We study a zero-sum stochastic game on a Borel state space where the state of the game is not known to the players. Both players take their decisions based on an observation process. We transform this into an equivalent problem with complete information. Then, we establish the existence of a value and optimal strategies for both players.     

Microeconomics-based resource allocation in overlay networks by using non-strategic behavior modeling

Behavior modeling has recently been investigated for designing self-organizing mechanisms in the context of communication networks in order to exploit the natural selfishness of the users with the goal of maximizing the overall utility. In strategic behavior modeling, the users of the network are assumed to be game players who seek to maximize their utility with taking into account the decisions that the other players might make. The essential difference between the aforementioned researches and this work is that it incorporates the non-strategic decisions in order to design the mechanism for the overlay network. In this solution concept, the decisions that a peer might make does not affect the actions of the other peers at all. The theory of consumer-firm developed in microeconomics is a model of the non-strategic behavior that we have adopted in our research. Based on it, we have presented distributed algorithms for peers’ “joining” and “leaving” operations. We have modeled the overlay network as a competitive economy in which the content provided by an origin server can be viewed as commodity and the origin server and the peers who multicast the content to their downside are considered as the firms. On the other hand, due to the dual role of the peers in the overlay network, they can be considered as the consumers as well. On joining to the overlay economy, each peer is provided with an income and tries to get hold of the service regardless to the behavior of the other peers. We have designed the scalable algorithms in such a way that the existence of equilibrium price (known as Walrasian equilibrium price) is guaranteed.     

Transparent prioritisation,budgeting and resource allocation with multi-criteria decision analysis and decision conferencing

Managers in both for-profit and not-for-profit organisations continually face the task of allocating resources by balancing costs, benefits and risks and gaining commitment by a wide constituency of stakeholders to those decisions. This task is complex and difficult because many options are present, benefits and risks are rarely expressed as single objectives, multiple stakeholders with different agendas compete for limited resources, individually optimal resource allocations to organisational units are rarely collectively optimal, and those dissatisfied with the decisions taken may resist implementation. We first explain three current approaches to resource allocation taken from corporate finance, operational research and decision analysis, and we identify a common mistake organisations make in allocating resources. The paper then presents a technical process, multi-criteria portfolio analysis, for balancing the conflicting elements of the problem, and a social process, decision conferencing, which engages all the key players during the modelling process, ensuring their ownership of the model and the subsequent implementation. This socio-technical process improves communication within the organisation, develops shared understanding of the portfolio and generates a sense of common purpose about those projects that will best realise the organisation’s objectives. The paper concludes with lessons we have learned from actual practice. The authors want to thank Allergan and FCT (Portuguese Science Foundation) for their support.     

The principal-agent problem with adaptive players

In this study we show predictions made by the standard principal-agent theory may not hold when knowlege assumptions are relaxed. Conventional principal-agent models assume players are completely rational: they know their own and other player's utilities and probabilities of all states of nature. In reality, players must make decisions without such knowledge.We define a simple version of the principal-agent game and examine it using object-oriented computer simulation. Player learning is modeled with a statistical learning model. Our results show that even this simple game combined with standard learning assumptions results in complex behavior. Expectations of both the principal and the agents are crucial in determining the system outcomes. Expectations and lack of prior knowledge make it possible for the principal to converge on suboptimal behavior or not converge on a consistent behavior at all. The same attributes in the agents make it possible for the principal to drive expectations down and thus get higher effort for lower reward.This study contributes a more robust understanding of the principal-agent model and its application to incentive design.     

Neural nets in a group decision process

In recent years there has been some interest in applying Artificial Adaptive Agents (AAA) to the study of complex adaptive systems, especially economic systems. Neural networks are frequently employed as AAA. Artificial neural nets mimic certain aspects of the physical structure and information processing of the human brain and their most attractive characteristic is their ability to learn a pattern from a given set of examples. In this study, we investigated the ability of neural nets to model human behavior in a group decision process. The context was a market entry game with a linear payoff function and binary decisions. The players had to decide, for each trial, whether or not to enter a market whose capacity is public knowledge. Human behavior in this situation has been modeled and empirically validated by the Nash equilibrium for noncooperative n-person games. A simulation of the game was performed with neural nets instead of human subjects. The nets were trained using the results of the games in which they participated. The simulation with groups of neural nets exhibits phenomena very similar to those observed in groups of human players. Received February 2000     

Static and Dynamic Equilibria in Games With Continuum of Players

This paper is a study of a general class of deterministic dynamic games with an atomless measure space of players and an arbitrary time space. The payoffs of the players depend on their own strategy, a trajectory of the system and a function with values being finite dimensional statistics of static profiles. The players' available decisions depend on trajectories of the system.The paper deals with relations between static and dynamic open-loop equilibria as well as their existence. An equivalence theorem is proven and theorems on the existence of a dynamic equilibrium are shown as consequences.     

Evolutionary insights on the willingness to communicate

While in previous models of pre-play communication players are forced to communicate, we investigate what happens if players can choose not to participate in this cheap talk. Outcomes are predicted by analyzing evolutionary stability in a population of a priori identical players. If the game following the communication rewards players who choose the same action then an efficient outcome is only guaranteed when participation in the pre-play communication is voluntary. If however players aim to coordinate on choosing different actions in the underlying game and there are sufficiently many messages then the highest payoff is selected when players are forced to talk to each other before playing the game. Received: September 2000/Revised: March 2003     

The Role of Emotions in Multi-Organizational Decision-Making

This paper addresses two questions. (i) What is the role of emotions in decision-making? (ii) Can emotions be attributed to organizations, and is their role in decisions made between organizations the same as in decisions made between individual people? The answers given to question (ii) are yes and yes. The proposed answer to question (i) is based upon a game-theoretic approach, which considers how emotions may be of strategic benefit to game players.     

Mediator: Towards a negotiation support system

Mediator is a negotiation support system (NSS) based on evolutionary systems design (ESD) and database-centered implementation. It supports negotiations by consensus seeking through exchange of information and, where consensus is incomplete, by compromise. The negotiation problem is shown — graphically or as relational data in matrix form — in three spaces as a mapping from control space to goal space and (through marginal utility functions) to utility space. Within each of these spaces the negotiation process is characterized by adaptive change, i.e., mappings of group target and feasible sets by which these sets are redefined in seeking a solution characterized by a single-point intersection between them.This concept is being implemented in Mediator, a data-based micro-mainframe NSS intended to support the players and a human mediator in multi-player decision situations. Each player employs private and shared database views, using his/her own micro-computer decision support system enhanced with a communications manager to interact with the mediator DSS. Sharing of views constitutes exchange of information which can lead towards consensus. The human mediator can support compromise, as needed, through use of solution concepts and/or concession-making procedures in the NSS model base. As a concrete example, we demonstrate the use of the system for group car buying decisions.     

Distributed consensus in noncooperative inventory games

This paper deals with repeated nonsymmetric congestion games in which the players cannot observe their payoffs at each stage. Examples of applications come from sharing facilities by multiple users. We show that these games present a unique Pareto optimal Nash equilibrium that dominates all other Nash equilibria and consequently it is also the social optimum among all equilibria, as it minimizes the sum of all the players’ costs. We assume that the players adopt a best response strategy. At each stage, they construct their belief concerning others probable behavior, and then, simultaneously make a decision by optimizing their payoff based on their beliefs. Within this context, we provide a consensus protocol that allows the convergence of the players’ strategies to the Pareto optimal Nash equilibrium. The protocol allows each player to construct its belief by exchanging only some aggregate but sufficient information with a restricted number of neighbor players. Such a networked information structure has the advantages of being scalable to systems with a large number of players and of reducing each player’s data exposure to the competitors.     

Implementability of correlated and communication equilibrium outcomes in incomplete information games

In a correlated equilibrium, the players’ choice of actions is directed by correlated random messages received from an outside source, or mechanism. These messages allow for more equilibrium outcomes than without any messages (pure-strategy equilibrium) or with statistically independent ones (mixed-strategy equilibrium). In an incomplete information game, the messages may also reflect the types of the players, either because they are affected by extraneous factors that also affect the types (correlated equilibrium) or because the players themselves report their types to the mechanism (communication equilibrium). Mechanisms may be further differentiated by the connections between the messages that the players receive and their own and the other players’ types, by whether the messages are statistically dependent or independent, and by whether they are random or deterministic. Consequently, whereas for complete information games there are only three classes of equilibrium outcomes, with incomplete information the corresponding number is 14 or 15 for correlated equilibria and even larger—15, 16 or 17—for communication equilibria. For both solution concepts, the implication relations between the different kinds of equilibria form a two-dimensional lattice, which is considerably more intricate than the single-dimensional one of the complete information case.     

On bargaining partner selection when communication is restricted

This paper analyzes the optimal selection of a bargaining partner when communication among players is restricted by an exogenously given graph. If players are equally patient, bargaining agreements are immune to players' locations, and the selection of a bargaining partner is not an issue. In contrast, when players differ in their discount factors, both location and impatience matter for bilateral agreements and partner selection becomes an issue. We show that selecting the most impatient neighbor is an equilibrium strategy whenever two players having a common neighbor share their most impatient neighbor. This condition is always satisfied by stratified graphs with no more than three strata. In the absence of this condition, cyclically-stratified graphs with no more than three strata also admit this equilibrium selection rule. Received: November 2000/Final version: January 2002     

Sharing beliefs about actions

We consider a strategic situation in which each player may not know the probability distribution governing the information structures of his opponents, and consequently his beliefs about opponents' action choices are represented by a set of probability measures. Suppose that beliefs of all the players are common knowledge. Then for any subset of players, the marginal beliefs of those players (about the action choices of their common opponents) must share at least one probability measure.     

非平衡拓扑和随机干扰情形下多自主体系统的趋同条件

研究了具有一般有向通信拓扑和高斯通信噪声的多自主体系统的趋同条件.这里所研究的有向拓扑不仅包含有向平衡图,而且包含非平衡图,后者是本文的重点.我们利用马氏链的结果得到了一个网络节点的互通类;通过对噪声影响的细化,给出了不同噪声情形下系统趋同条件:(1)对互通类的自主体获取信息受到噪声干扰情形,给出了系统均方趋同的充要条件,并证明该条件也保证以概率1 趋同;(2)对互通类的自主体获取信息未受到噪声干扰但其余自主体获取信息受到干扰情形,给出了系统均方趋同的充分条件,并证明该条件在一定意义下也是必要的;(3)对整个系统无噪声情形,给出了系统趋同的充要条件.     

A two-stage supply chain coordination mechanism considering price sensitive demand and quantity discounts

This paper explores the coordination between a supplier and a buyer within a decentralized supply chain, through the use of quantity discounts in a game theoretic model. Within this model, the players face inventory and pricing decisions. We propose both cooperative and non-cooperative approaches considering that the product traded experiences a price sensitive demand. In the first case, we study the dynamics of the game from the supplier's side as the leader in the negotiation obtaining a Stackelberg equilibrium, and then show how the payoff of this player could still improve from this point. In the second case, a cooperative model is formulated, where decisions are taken simultaneously, emulating a centralized firm, showing the benefits of the cooperation between the players. We further formulate a pricing game, where the buyer is allowed to set different prices to the final customer as a reaction to the supplier's discount decisions. For the latter we investigate the difference between feasibility of implementing a retail discount given a current coordination mechanism and without it. Finally the implications of transportation costs are analyzed in the quantity discount schedule. Our findings are illustrated with a numerical example showing the difference in the players’ payoff in each case and the optimal strategies, comparing in each case our results with existing work.     

无中心优化的算子分裂方法

在某些多智能体系统中,由于受到通讯等因素的限制,单个智能体只能进行本地计算,再与相邻智能体交换数据.与传统的并行和分布式计算不同,这种数据交换方式不再使用中心节点或者共享内存,而仅限于相邻节点之间.这种通过局部数据交换而实现全网目标的方式叫做无中心计算.比如,从任意的多个数开始,所有智能体通过不断地计算其局部平均,就都能收敛到这些数的平均值.无中心计算有不易形成通讯和计算瓶颈的优点,更适合分布的节点,因此受到一些应用的欢迎. 本文介绍求解一致最优化问题的若干无中心算法.一致最优化问题的目标是全网所有节点的变量收敛到同一个、并使所有目标函数之和最小的值.我们可以通过推广求平均的无中心方法去实现这个目标,但是得到算法比普通（有中心的）优化算法收敛得更慢,有阶数差距.近年来,一些新的无中心算法弥补了这个阶数差距.本文采用算子分裂的统一框架,以比这些算法原文更为简单的形式介绍这些方法.     

A two-player competitive discrete location model with simultaneous decisions

In this paper we will describe and study a competitive discrete location problem in which two decision-makers (players) will have to decide where to locate their own facilities, and customers will be assigned to the closest open facilities. We will consider the situation in which the players must decide simultaneously, unsure about the decisions of one another, and we will present the problem in a franchising environment. Most problems described in the literature consider sequential rather than simultaneous decisions. In a competitive environment, most problems consider that there is a set of known and already located facilities, and new facilities will have to be located, competing with the existing ones. In the presence of more than one decision-maker, most problems found in the literature belong to the class of Stackelberg location problems, where one decision-maker, the leader, locates first and then the other decision-maker, the follower, locates second, knowing the decisions made by the first. These types of problems are sequential and differ significantly from the problem tackled in this paper, where we explicitly consider simultaneous, non-cooperative discrete location decisions. We describe the problem and its context, propose some mathematical formulations and present an algorithmic approach that was developed to find Nash equilibria. Some computational tests were performed that allowed us to better understand some of the features of the problem and the associated Nash equilibria. Among other results, we conclude that worsening the situation of a player tends to benefit the other player, and that the inefficiency of Nash equilibria tends to increase with the level of competition.     

Operational Research for Complex Conflicts: A Gaming Methodology for the Development of a Decision-Making Monitor

A methodology for providing assistance to decision-makers in conflicts is demonstrated. It is shown that in simple "Chicken" games, conflict-like situations arise only when players have radically different aims and that players experiencing crisis-like situations in such games typically exhibit behaviours which would exacerbate real conflicts. It is argued that explicit knowledge of such behavioural tendencies. the conditions in which they arise, and the possible underlying mechanisms, would be of assistance to decision-makers, both for them to monitor their own behaviour, and to assess that of other parties.