Peer-to-Peer (P2P) networks have seen tremendous growth in development and usage in recent times. This attention has brought many developments as well as new challenges to these networks. We will show that agent extensions to P2P networks offer solutions to many problems faced by P2P networks. In this research, an attempt is made to bring together JXTA P2P infrastructure and Jinni, a Prolog based agent engine to form an agent based P2P network. On top of the JXTA, we define simple Java API providing P2P services for agent programming constructs. Jinni is deployed on this JXTA network using an automated code update mechanism. Experiments are conducted on this Jinni/JXTA platform to implement a simple agent communication and data exchange protocol.

D.Phil. / A modern day N-P complete problem is the assigning of frequencies to transmitters in a cellular network in such a manner that, ideally, no two transmitters in the same cell or neighbouring cells use the same frequency. Considering that an average cellular network provider has over 29 000 transmitters and only 55 frequencies, choosing these frequencies in an optimal way is a very difficult computational problem. Swarm intelligence allows the acceptable minimization and optimization of the frequency assignment problem (FAP). Swarm intelligence is a concept modelling the processes in natural systems such as ant colonies, beehives, human immune systems and the human brain. These systems are selforganizational and display high efficiency in the execution of their tasks. A number of simple automated agents interacting with each other and the environment form a collective. Specifically, there is no "central agent" directing the others. A collective can display surprising intelligence which emerges out of the interaction of the individual agents. This collective intelligence, referred to as swarm intelligence, is displayed in ant colonies when ants build elaborate nests, regulate nest temperature and efficiently search for food in very complex environments. In this thesis a proposal is made to utilize swarm intelligence to build a swarm automatic frequency planner (swarm AFP). The swarm AFP produces frequency plans that are better, or on par with existing frequency planning tools, and in a fraction of the time. A swarm AFP is presented through an in-depth investigation into complex adaptive systems, agent architectures and emergence. Based on an understanding of these concepts, a swarm intelligence model called ACEUS is constructed. ACEUS forms the platform of the swarm AFP. It is a contribution to multi-agent technology as it is a new multi-agent framework that exhibits swarm intelligence and complex distributed computation. What differentiates ACEUS from other multi-agent technologies is that ACEUS works on the basis that the tasks or constructions that have been created by the agents actually guide the agents in their endeavours. There is no centralised agent controlling or guiding the process. The agents in ACEUS receive information and stimulation from their tasks or constructions in the environment. As these constructions or tasks alter the environment, the agents receive stimulus from the changing environment and then react to the changing environment. The changing environment acts as an emergent guiding force to the agents. This is the important contribution that stigmergy contributes to ACEUS. Utilizing this concept, ACEUS is used to create a swarm AFP. The swarm AFP is benchmarked against the COST 259 Siemens benchmarks. In all the COST 259 Siemens scenarios the swarm AFP produced the best results in the shortest time. The swarm AFP was also tested in a real cellular network and the resulting statistics before and after the swarm AFP implementation are presented.

The increasingly popular cloud-computing paradigm provides on-demand access to computing and storage with the appearance of unlimited resources. Users are given access to a variety of data and software utilities to manage their work. Users rent virtual resources and pay for only what they use. In spite of the many benefits that cloud computing promises, the lack of dependability in shared virtualized infrastructures is a major obstacle for its wider adoption, especially for mission-critical applications. Virtualization and multi-tenancy increase system complexity and dynamicity. They introduce new sources of failure degrading the dependability of cloud computing systems. To assure cloud dependability, in my dissertation research, I develop autonomic failure identification and diagnosis techniques that are crucial for understanding emergent, cloud-wide phenomena and self-managing resource burdens for cloud availability and productivity enhancement. We study the runtime cloud performance data collected from a cloud test-bed and by using traces from production cloud systems. We define cloud signatures including those metrics that are most relevant to failure instances. We exploit profiled cloud performance data in both time and frequency domain to identify anomalous cloud behaviors and leverage cloud metric subspace analysis to automate the diagnosis of observed failures. We implement a prototype of the anomaly identification system and conduct the experiments in an on-campus cloud computing test-bed and by using the Google datacenter traces. Our experimental results show that our proposed anomaly detection mechanism can achieve 93% detection sensitivity while keeping the false positive rate as low as 6.1% and outperform other tested anomaly detection schemes. In addition, the anomaly detector adapts itself by recursively learning from these newly verified detection results to refine future detection.

As bandwidth constraints on LAN/WAN environments decrease, the demand for distributed services will continue to increase. In particular, the proliferation of user-level applications requiring high-capacity distributed file storage systems will demand that such services be universally available. At the same time, the advent of high-speed networks have made the deployment of application and communication solutions based upon an Intelligent Mobile Agent (IMA) framework practical. Agents have proven to present an ideal development paradigm for the creation of autonomous large-scale distributed systems, and an agent-based communication scheme would facilitate the creation of independently administered distributed file services. This thesis thus outlines an architecture for such a distributed file system based upon an IMA communication framework.

University of Macau / Faculty of Science and Technology. / Department of Computer and Information Science

本論文包含四個部分。每一部分我們演示一個在共同演化網絡中的個體為本(agent-based)模型。第二章是不滿適應雪堆博奕(DASG)的廣泛化。第三章是自省適應(self-questioning adaptive)雪堆博奕。第四章是共同演化選民模型的廣泛化。第五章是有三個互相克制的角色的適應性石頭-布-剪刀(ARPS)模型。在這些模型中，適應行為導致共同演化過程發生。我們以電算模擬及理論方法研究這些模型。我們的目標是建立一個可應用於不同共同演化網絡的一般分析框架。 / 在第二章及第四章，我們將Gräser等人的DASG及Vazquez等人的共同演化選民模型從一個控制參數推廣到二個獨立的控制參數。在他們的工作中，根據網絡的結構定義了一些相，而且發展了平均場理論。而在廣泛化的情況下，在已伸延的相空間上，我們也定義了一些相及發展了一些廣泛化的平均場理論。在廣泛化DASG中，我們以考慮在相邊界附近的最終生存形態(last surviving patterns)以解釋相邊界的電算模擬結果。 / 在第三章，我們提出及研究一個以誘惑驅動的雪堆博奕。該更新機制被稱為自省機制(self-questioning mechanism)。我們給出模擬及理論結果，也討論了該些結果的物理意義。 / 在第五章，我們推廣我們的研究至有三個策略的遊戲。我們提出及研究了一個ARPS模型，其中每個個體採用三個互相克制的策略的其中之一。每個個體以概率 p來重連不理想的連結或以概率 (1 - p)改變自身的策略以適應其周遭環境。我們研究了網絡於不同的 p值在穩定態的行為及定義 了一些相。我們研究兩個選取重連對象的方法，分別對應於隨機選取及刻意選取重連對象，也解釋了得出的結果。我們在有關穩定勝利、平手及失敗概率的研究中及哪種個體可以有更高的勝利概率的研究中得出了有趣的結果。我們也研究了結果如何取決於初始條件。 / 在不同的模型中，理論方程均建立於相似的想法上。理論結果得出模擬結果的主要特性，包括出現了不同的相。該分析方法被證明了在本論文中對不同的模型也有效，而該方法也可被應用於很多其他共同演化網絡上。 / This thesis consists of four parts. In each part, we present results of an agent-based model of co-evolving network. Chapter 2 deals with a generalization of the Dissatisfied Adaptive Snowdrift Game (DASG) and Chapter 3 covers the self-questioning adaptive snowdrift game. Chapter 4 discusses a generalization of a co-evolving voter model. Chapter 5 gives the results on a cyclic three-character Adaptive Rock-Paper-Scissor (ARPS) game. The adaptive actions give rise to co-evolving processes in these models. These models are studied both numerically and analytically. An objective here is to establish a general analytic framework that is applicable to different models of co-evolving networks. / In Chapters 2 and 4, we generalize two existing models -the DASG of Gräser et al. and the co-evolving voter model of Vazquez et al. -from a single control parameter to two independent parameters. Different phases were identified according to the network structure and mean-field theories were developed in the previous work. With the expanded phase space in our generalized models, we identified different phases and provided a generalized mean-field approach. The phase boundaries in the generalized DASG can be explained by considering the last surviving patterns in the vicinity of the transition between two phases. / In Chapter 3, we propose and study a co-evolving snowdrift game in which the adaptive actions are driven by the desire to enhance winning. The updating scheme is called the self-questioning mechanism. We present simulation and theoretical results, and provide physical meaning to the results. / In Chapter 5, we extend our study to three-strategy games. An ARPS model in which each agent uses one of three strategies that dominate each other cyclically is proposed and studied. Each agent adapts his local environment by rewiring an un-favourable link with a probability p or switching his strategy with a probability 1-p. As p varies, the behaviour of the network in the steady state is studied and dierent phases are identified. Two dierent schemes corresponding to selecting the rewiring target randomly and intentionally are studied and the results are explained. Interesting results are also found in the probabilities of winning, losing and drawing in the steady state; and the type of agents that have a higher winning probability. The dependence on the initial distribution of the three strategies among the agents is also studied. / The analytic equations presented for each model in the thesis are established on similar ideas. The analytic results capture the main features in the simulation results, including the emergence of dierent phases. The analytic approach, proven to be useful through different models in this thesis, can be applied to a wide class of other co-evolving network models. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Choi, Chi Wun / 個體為本共同演化網絡的研究 / 蔡至桓. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 114-116). / Abstracts also in Chinese. / Choi, Chi Wun / Ge ti wei ben gong tong yan hua wang luo de yan jiu / Cai Zhihuan. / Abstract --- p.i / 摘要 --- p.iii / Acknowledgements --- p.v / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Introduction --- p.1 / Chapter 1.2 --- Review --- p.5 / Chapter 1.2.1 --- Network and basic graph properties --- p.5 / Chapter 1.2.2 --- Two-person games --- p.6 / Chapter 2 --- Generalization of Dissatisfied-Adaptive Snowdrift Game (DASG) --- p.8 / Chapter 2.1 --- Introduction --- p.8 / Chapter 2.2 --- Dissatisfied-Adaptive model --- p.12 / Chapter 2.3 --- Previous work --- p.14 / Chapter 2.4 --- Generalized Dissatisfied-Adaptive model --- p.16 / Chapter 2.5 --- Simulation results --- p.17 / Chapter 2.6 --- Theoretical analysis --- p.19 / Chapter 2.6.1 --- Mean-Field approach --- p.19 / Chapter 2.6.2 --- Theoretical results --- p.22 / Chapter 2.7 --- Last surviving patterns --- p.25 / Chapter 2.7.1 --- Observing the last surviving patterns --- p.25 / Chapter 2.7.2 --- Applying the theory using extracted information from simulations --- p.26 / Chapter 2.7.3 --- Further development of the theory --- p.28 / Chapter 2.7.4 --- Results of the theory --- p.30 / Chapter 2.8 --- Dependence on initial conditions and mean degree --- p.32 / Chapter 2.9 --- Conclusion --- p.34 / Chapter 3 --- Self-questioning Adaptive SG --- p.36 / Chapter 3.1 --- Introduction --- p.36 / Chapter 3.2 --- Self-questioning adaptive SG with control parameter r --- p.39 / Chapter 3.2.1 --- Model --- p.39 / Chapter 3.2.2 --- Results --- p.40 / Chapter 3.3 --- Self-questioning adaptive SG with control parameters r and h --- p.42 / Chapter 3.3.1 --- Model --- p.42 / Chapter 3.3.2 --- Results --- p.43 / Chapter 3.4 --- Conclusion --- p.45 / Chapter 4 --- Generalization of co-evolving voter model --- p.46 / Chapter 4.1 --- Introduction --- p.46 / Chapter 4.2 --- Co-evolving voter model --- p.49 / Chapter 4.3 --- Previous work --- p.50 / Chapter 4.4 --- Simulation results --- p.52 / Chapter 4.4.1 --- Results of macroscopic quantities --- p.52 / Chapter 4.4.2 --- Results of trajectories by simulations --- p.54 / Chapter 4.4.3 --- The largest component --- p.55 / Chapter 4.4.4 --- Short Summary --- p.56 / Chapter 4.5 --- Theoretical analysis --- p.57 / Chapter 4.5.1 --- Mean-Field approach --- p.57 / Chapter 4.5.2 --- Theoretical results --- p.59 / Chapter 4.6 --- Dependence on initial conditions and mean degree --- p.60 / Chapter 4.6.1 --- Results for different mean degrees --- p.60 / Chapter 4.6.2 --- Results for different initial conditions --- p.61 / Chapter 4.7 --- Conclusion --- p.63 / Chapter 5 --- Adaptive Rock-Paper-Scissors games --- p.64 / Chapter 5.1 --- Introduction --- p.64 / Chapter 5.2 --- Adaptive Rock-Paper-Scissors Model --- p.67 / Chapter 5.3 --- Simulation results --- p.70 / Chapter 5.4 --- Theoretical analysis --- p.73 / Chapter 5.4.1 --- Simplifications by threefold-symmetry --- p.73 / Chapter 5.4.2 --- Changes in local quantities --- p.74 / Chapter 5.4.3 --- Mean-Field approach --- p.75 / Chapter 5.4.4 --- Theoretical results --- p.80 / Chapter 5.5 --- Dependence on mean degree --- p.82 / Chapter 5.6 --- Oriented rewiring method --- p.83 / Chapter 5.7 --- Probabilities of winning, drawing and losing --- p.85 / Chapter 5.7.1 --- Average probabilities of winning, drawing and losing in the steady state --- p.85 / Chapter 5.7.2 --- Degree distribution and the distributions of the probabilities --- p.86 / Chapter 5.7.3 --- Brief explanation --- p.88 / Chapter 5.7.4 --- Results for a larger μ --- p.89 / Chapter 5.7.5 --- Short summary --- p.90 / Chapter 5.8 --- Results for general initial conditions --- p.92 / Chapter 5.8.1 --- Coupled dynamical equations --- p.92 / Chapter 5.8.2 --- Trajectories of the macroscopic quantities --- p.94 / Chapter 5.8.3 --- Phases and theoretical ternary phase diagrams --- p.96 / Chapter 5.9 --- Conclusion --- p.98 / Chapter 6 --- Summary --- p.100 / Chapter A --- Coupled dynamical equations for Self-questioning adaptive SG --- p.104 / Chapter B --- Theoretical results for Self-questioning adaptive SG with control parameters r and h --- p.106 / Chapter C --- Derivations of Mean-Field equations in ARPS model --- p.108 / Chapter D --- Derivations of Mean-Field equations for the oriented rewiring method in ARPS model --- p.111 / Bibliography --- p.114

by Zhao Lei. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2002. / Includes bibliographical references (leaves 78-85). / Abstracts in English and Chinese. / Abstract --- p.ii / Acknowledgments --- p.iv / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Background --- p.5 / Chapter 2.1 --- Agents --- p.5 / Chapter 2.1.1 --- Deliberative Agents --- p.7 / Chapter 2.1.2 --- Reactive Agents --- p.8 / Chapter 2.1.3 --- Interacting Agents --- p.9 / Chapter 2.1.4 --- Hybrid Architectures --- p.10 / Chapter 2.2 --- Real-time Artificial Intelligence --- p.10 / Chapter 2.3 --- Real-Time Agents --- p.12 / Chapter 2.3.1 --- The Subsumption Architecture --- p.13 / Chapter 2.3.2 --- The InterRAP Architecture --- p.15 / Chapter 2.3.3 --- The 3T Architecture --- p.16 / Chapter 2.4 --- On-line Scheduling in Real-Time Agents --- p.18 / Chapter 3 --- A Real-Time Agent Architecture --- p.20 / Chapter 3.1 --- Human Cognition Model --- p.20 / Chapter 3.1.1 --- Perception --- p.22 / Chapter 3.1.2 --- Cognition --- p.22 / Chapter 3.1.3 --- Action --- p.23 / Chapter 3.2 --- Real-Time Message Passing Primitives and Process Structuring --- p.24 / Chapter 3.2.1 --- Message Passing as IPC --- p.25 / Chapter 3.2.2 --- Administrator and Worker Processes --- p.28 / Chapter 3.3 --- Agent Architecture --- p.29 / Chapter 3.3.1 --- Sensor Workers and the Sensor Administrator --- p.30 / Chapter 3.3.2 --- The Cognition Workers --- p.32 / Chapter 3.3.3 --- "The Task Administrator, the Scheduler Worker and Ex- ecutor Workers" --- p.32 / Chapter 3.4 --- An Agent-Based Real-time Arcade Game --- p.34 / Chapter 4 --- A Multiple Method Approach to Task Scheduling --- p.37 / Chapter 4.1 --- Task Scheduling Mechanism --- p.37 / Chapter 4.1.1 --- Task and Action --- p.38 / Chapter 4.1.2 --- Task Administrator --- p.40 / Chapter 4.1.3 --- Task Scheduler --- p.43 / Chapter 4.2 --- A Task Scheduling Model --- p.44 / Chapter 4.3 --- Combination Rules and Special Cases --- p.46 / Chapter 4.4 --- Scheduling Algorithms --- p.49 / Chapter 5 --- Task Scheduling Model: Analysis and Experiments --- p.53 / Chapter 5.1 --- Goodness Measure --- p.53 / Chapter 5.2 --- Theoretical Analysis --- p.54 / Chapter 5.3 --- Implementation --- p.59 / Chapter 5.3.1 --- Task Generator Implementation --- p.59 / Chapter 5.3.2 --- Executor Workers Implementation --- p.61 / Chapter 5.4 --- Experimental Results --- p.62 / Chapter 5.4.1 --- Hybrid Mechanism and Individual Algorithms --- p.63 / Chapter 5.4.2 --- Effect of Average Execution Time --- p.65 / Chapter 5.4.3 --- Effect of the Greedy Algorithm --- p.65 / Chapter 5.4.4 --- Effect of the Advanced Algorithm --- p.67 / Chapter 5.4.5 --- Effect of Actions and Relations Among Them --- p.68 / Chapter 5.4.6 --- Effect of Deadline --- p.71 / Chapter 6 --- Conclusions --- p.73 / Chapter 6.1 --- Summary of Contributions --- p.73 / Chapter 6.2 --- Future Work --- p.75

Lee Ting-on. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2000. / Includes bibliographical references (leaves 116-121). / Abstracts in English and Chinese. / Acknowledgments --- p.i / Abstract --- p.ii / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Roadmap to the Thesis --- p.5 / Chapter 2 --- Software Agents and Agent Frameworks --- p.7 / Chapter 2.1 --- Software Agent --- p.7 / Chapter 2.1.1 --- Advantages of Agent --- p.10 / Chapter 2.1.2 --- Roles of Agent --- p.11 / Chapter 2.2 --- Agent Frameworks --- p.13 / Chapter 2.3 --- Communication Services and Concepts --- p.15 / Chapter 2.3.1 --- Message Channel --- p.15 / Chapter 2.3.2 --- Remote Procedure Call --- p.16 / Chapter 2.3.3 --- Event Channel --- p.17 / Chapter 2.4 --- Component --- p.18 / Chapter 3 --- Related Work --- p.20 / Chapter 3.1 --- Collaboration Behaviors --- p.20 / Chapter 3.2 --- Direct Coordination --- p.22 / Chapter 3.3 --- Meeting-oriented Coordination --- p.23 / Chapter 3.4 --- Blackboard-based Coordination --- p.24 / Chapter 3.5 --- Linda-like Coordination --- p.25 / Chapter 3.6 --- Reactive Tuple Spaces --- p.26 / Chapter 4 --- Background and Foundations --- p.27 / Chapter 4.1 --- Choice of Technologies --- p.27 / Chapter 4.2 --- Jini Technology --- p.28 / Chapter 4.2.1 --- The Lookup Service --- p.29 / Chapter 4.2.2 --- Proxy --- p.31 / Chapter 4.3 --- JavaSpaces --- p.32 / Chapter 4.4 --- Grasshopper Architecture --- p.33 / Chapter 5 --- The CoDAC Framework --- p.36 / Chapter 5.1 --- Requirements for Enabling Collaboration --- p.37 / Chapter 5.1.1 --- Consistent Group Membership --- p.37 / Chapter 5.1.2 --- Atomic Commitment --- p.39 / Chapter 5.1.3 --- Uniform Reliable Multicast --- p.40 / Chapter 5.1.4 --- Fault Tolerance --- p.40 / Chapter 5.2 --- System Components --- p.41 / Chapter 5.2.1 --- Distributed Agent Adapter --- p.42 / Chapter 5.2.2 --- CollaborationCore --- p.44 / Chapter 5.3 --- System Infrastructure --- p.45 / Chapter 5.3.1 --- Agent --- p.45 / Chapter 5.3.2 --- Distributed Agent Manager --- p.46 / Chapter 5.3.3 --- Collaboration Manager --- p.46 / Chapter 5.3.4 --- Kernel --- p.46 / Chapter 5.4 --- Collaboration --- p.47 / Chapter 5.5.1 --- Global Collaboration --- p.48 / Chapter 5.5.2 --- Local Collaboration --- p.48 / Chapter 6 --- Collaboration Life Cycle --- p.50 / Chapter 6.1 --- Initialization --- p.50 / Chapter 6.2 --- Resouces Gathering --- p.53 / Chapter 6.3 --- Results Delivery --- p.54 / Chapter 7 --- Protocol Suite --- p.55 / Chapter 7.1 --- The Group Membership Protocol --- p.56 / Chapter 7.1.1 --- Join Protocol --- p.56 / Chapter 7.1.2 --- Leave Protocol --- p.57 / Chapter 7.1.3 --- Recovery Protocol --- p.59 / Chapter 7.1.4 --- Proof --- p.61 / Chapter 7.2 --- Atomic Commitment Protocol --- p.62 / Chapter 7.3 --- Uniform Reliable Multicast --- p.63 / Chapter Chapter 8 --- Implementation --- p.66 / Chapter 8.1 --- Interfaces and Classes --- p.66 / Chapter 8.1.1 --- The CoDACAdapterInterface --- p.66 / Chapter 8.1.2 --- The CoDACEventListener --- p.69 / Chapter 8.1.3 --- The DAAdapter --- p.71 / Chapter 8.1.4 --- The DAManager --- p.75 / Chapter 8.1.5 --- The CoDACInternalEventListener --- p.77 / Chapter 8.1.6 --- The CollaborationManager --- p.77 / Chapter 8.1.7 --- The CollaborationCore --- p.78 / Chapter 8.2 --- Messaging Mechanism --- p.79 / Chapter 8.3 --- Nested Transaction --- p.84 / Chapter 8.4 --- Fault Detection --- p.85 / Chapter 8.5 --- Atomic Commitment Protocol --- p.88 / Chapter 8.5.1 --- Message Flow --- p.89 / Chapter 8.5.2 --- Timeout Actions --- p.91 / Chapter Chapter 9 --- Example --- p.93 / Chapter 9.1 --- System Model --- p.93 / Chapter 9.2 --- Auction Lifecycle --- p.94 / Chapter 9.2.1 --- Initialization --- p.94 / Chapter 9.2.2 --- Resource Gathering --- p.98 / Chapter 9.2.3 --- Results Delivery --- p.100 / Chapter Chapter 10 --- Discussions --- p.104 / Chapter 10.1 --- Compatibility --- p.104 / Chapter 10.2 --- Hierarchical Group Infrastructure --- p.106 / Chapter 10.3 --- Flexibility --- p.107 / Chapter 10.4 --- Atomicity --- p.108 / Chapter 10.5 --- Fault Tolerance --- p.109 / Chapter Chapter 11 --- Conclusion and Future Work --- p.111 / Chapter 11.1 --- Conclusion --- p.111 / Chapter 11.2 --- Future Work --- p.112 / Chapter 11.2.1 --- Electronic Commerce --- p.112 / Chapter 11.2.2 --- Workflow Management --- p.114 / Bibliography --- p.116 / Publication List --- p.121

In infinitely repeated games, we also give definitions to risk attitude and reputation. As art infinitely repeated game is a repetition of a constituent strategic game, we transform each single round of an infinitely repeated game as a risk game. We extend the definitions of risk strategies and risk strategy equilibrium to infinitely repeated games. We also research some properties of risk strategy equilibrium and show that players can obtain higher payoffs in risk strategy equilibrium than in pure strategy Nash equilibrium. / In multi-agent systems, agents often need to make decisions, especially under uncertainty. When a decision-maker needs to choose among a number of choices, each having a certain probability to happen, one of the traditional ways discussed in economics is to calculate the expected utility of each choice and choose the one with the maximum expected utility. However, most of the humans do not do so in real situations. Very often, humans choose a choice with a lower expected utility. One of the famous examples is the Allais paradox. / In strategic games, we define risk attitude and reputation, which are factors that decision-makers take into account in making decisions. We transform a strategic game to a risk game. We propose a new kind of strategies, called risk strategies. In the transformed risk game, we find a new kind of equilibrium, called risk strategy equilibrium. We also find out some properties of risk strategy equilibrium. In addition, we find that players can obtain higher payoffs in risk strategy equilibrium than in pure strategy Nash equilibrium. / One of the key properties defining an intelligent agent is social ability. This means that an intelligent agent should be able to interact with other agents or humans. Before designing an intelligent agent for any multi-agent system, we need to first understand how agents should behave and interact in that particular application. / One way to understand how agents should behave in a particular application is to model the application as a game. Besides, many real-life situations can be modeled as games. So, we extend the model of attractiveness maximization and apply the extended model to strategic games and infinitely repeated games. / The reason why most of the people do not maximize the expected utility is that people have different attitudes towards risk in different situations and people are generally risk-averse. To model this human behavior, we propose another way of decision-making, called attractiveness maximization. In this model, every choice has an attractiveness, which is calculated from the risk attitude of the decision-maker, probability, and utility. In making decisions, decision-makers choose the choice with the maximum attractiveness. Using attractiveness maximization, the phenomenon that human do not maximize expected utility can be explained. We also find some properties of the model of attractiveness maximization, which match the human behaviors. / We develop several applications of attractiveness maximization and risk strategies. First, we apply the proposed concepts to the Iterated Prisoner's Dilemma, which is widely used by economists and sociologists to model and simulate many of the human interactions. Simulation shows that agents have improved performance and are reactive as well as pro-active. Second, we construct behavioral predictors and an adaptive strategy for Minority Games, which model many real life situations like the financial market, auctions and resources competitions. Simulations show that the adaptive strategy works much better than previous models. Third, we model a resource allocation problem as a Minority Game and apply the behavioral predictors and the adaptive strategy to the resource allocation problem. Simulations also show that agents with the proposed adaptive strategy are able to make more right decisions and better resource utilization than previous work. / Lam, Ka Man. / "August 2007." / Adviser: Leung Ho Fung. / Source: Dissertation Abstracts International, Volume: 69-02, Section: B, page: 1107. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2007. / Includes bibliographical references (p. 159-173). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract in English and Chinese. / School code: 1307.

Abstract not available