Massey Documents by Type

Permanent URI for this communityhttps://mro.massey.ac.nz/handle/10179/294

Browse

Has files: YesSubject: search.filters.subject.Intelligent agents (Computer software)

Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    Decision markets implementations for human forecasters and multi-agent learning systems : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Computer Science at Massey University, Albany, New Zealand
    (Massey University, 2023) Wang, Wenlong
    Mechanisms of collective decision-making are an increasingly important topic, given that relevant data and information are often distributed. Collective decision-making processes involve eliciting information from multiple agents, aggregating the information, and mapping the aggregated information to a decision. An obstacle to these processes is that information is often proprietary, held by self-interested agents, and sometimes even too sensitive to share. Decision markets are mechanisms for eliciting and aggregating such information into predictions for decision-making. A design for decision markets put forward by Chen, Kash, Ruberry, et al. uses prediction markets to elicit and aggregate predictions that are conditional to the available actions, and then uses a stochastic decision rule to determine, based on the aggregated forecasts, which action to select. The design is incentive-compatible and uses a decision scoring rule to evaluate and incentivise the self-interested agents for their forecasts. The first part of this thesis (Chapter 2) describes a framework for security-based decision markets that allows agents to make predictions by trading assets. Security-based decision markets are designed to be user-friendly for participants familiar with trading in stock markets. For prediction markets, such a framework is well studied. For decision markets, my results show there are important differences between scoring rule based and securities-based implementation. The second and third parts of this thesis (Chapters 3 and 4) investigate decision markets as mechanisms of collective decision-making for multi-agent learning problems, thus building a bridge between economic mechanisms and artificial intelligence. Chapter 3 provides a decision market based algorithm that allows a principal to train multiple autonomous agents with independent and identically distributed (iid) information to solve a contextual bandit problem. Simulation results demonstrate that the proposed multi-agent systems can achieve performance equivalent to a centralised counterpart without requiring direct access to the agents' iid information, which is necessary for the centralised counterpart. Chapter 4 describes a set of mechanisms that allow avoiding stochastic decision rules to select actions based on aggregated forecasts. This is important because committing to a stochastic (i.e., randomising) decision rule means that sometimes suboptimal decisions have to be taken. The mechanisms outlined in this chapter require agents to collectively predict a proxy instead of conditional outcomes. Simulations show that the performance is as good as a Bayesian model with access to all distributed information.
  • Thumbnail Image
    Item
    Synthesized cooperative strategies for intelligent multi-robots in a real-time distributed environment : a thesis presented in partial fulfillment of the requirements for the degree of Master of Science in Computer Science at Massey University, Albany, New Zealand
    (Massey University, 2009) Lin, Caoyun
    In the robot soccer domain, real-time response usually curtails the development of more complex Al-based game strategies, path-planning and team cooperation between intelligent agents. In light of this problem, distributing computationally intensive algorithms between several machines to control, coordinate and dynamically assign roles to a team of robots, and allowing them to communicate via a network gives rise to real-time cooperation in a multi-robotic team. This research presents a myriad of algorithms tested on a distributed system platform that allows for cooperating multi- agents in a dynamic environment. The test bed is an extension of a popular robot simulation system in the public domain developed at Carnegie Mellon University, known as TeamBots. A low-level real-time network game protocol using TCP/IP and UDP were incorporated to allow for a conglomeration of multi-agent to communicate and work cohesively as a team. Intelligent agents were defined to take on roles such as game coach agent, vision agent, and soccer player agents. Further, team cooperation is demonstrated by integrating a real-time fuzzy logic-based ball-passing algorithm and a fuzzy logic algorithm for path planning. Keywords Artificial Intelligence, Ball Passing, the coaching system, Collaborative, Distributed Multi-Agent, Fuzzy Logic, Role Assignment
  • Thumbnail Image
    Item
    Action-selection in RoboCup keepaway soccer : experimenting with player confidence : a thesis presented in partial fulfilment of the requirements for the degree of Masters of Science in Computer Science at Massey University
    (Massey University, 2006) Neilson, Samara Ann
    Through the investigation of collaborative multi-agent domains, in particular those of robot soccer and robot rescue, and the examination of many popular action-selection methodologies, this study identifies some of the issues surrounding entropy, action-selection and performance analysis. In order to address these issues, a meaningful method of on-field player evaluation, the confidence model, was first proposed then implemented as an action-selection policy. This model represented player skill through the use of percentages signifying relative strength and weakness and was implemented using a combination of ideas taken from Bayesian Theory. Neural Networks. Reinforcement Learning, Q-Learning and Potential Fields. Through the course of this study, the proposed confidence model action-selection methodology was thoroughly tested using the Keepaway Soccer Framework developed by Stone, Kuhlmann, Taylor and Liu and compared with the performance of its peers. Empirical test results were also presented, demonstrating both the viability and flexibility of this approach as a sound, homogeneous solution, for a team wishing to implement a quickly trainable performance analysis solution.
  • Thumbnail Image
    Item
    Two agent-based models of trust in social networks : a thesis presented in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Psychology
    (Massey University, 2008) Street, Susan Elizabeth
    Trust is a pervasive feature of human social interaction. Much of the recent interest in trust has been at the level of individuals and dyads. But trust is also important in networks, as it enables the formation and maintenance of social cooperation. Understanding this requires an understanding of how trust arises, functions, and is maintained within networks of people. Developing understandings of how individual behaviours aggregate, and how they evolve within an environment that includes other individuals developing similar behaviours is a difficult task. One way that it may be approached is through computer simulation using agent-based models. This thesis describes the development of two agent-based models of trust. Agent-based modelling is a novel method within the discipline of social psychology. The thesis first describes what agent-based modelling is, describes some of the situations in which it might be applicable, discusses how it might apply to modelling individuals in a social setting, and discusses the experience of developing the model. The first model was based on a theoretical cognitive model of behaviour within a particular formal game that has been claimed to involve trust, the Investor Game. This model showed that a population in which all individuals are are pursuing similar optimal strategies does not generate any of the interesting behaviours that we would expect to see in real-world interactions involving trust and cooperation. This tends to suggest that modelling trust behaviours also requires modelling behaviours that are untrustworthy, and representing a full range of potential behaviours, including outliers. The second model is based on a more naturalistic setting, on-line peer-to-peer trading through sites such as New Zealand's Trade Me, or eBay. In this model, individual traders carry characteristics that determine their reliability and honesty, and attempt to find effective strategies for identifying other traders' trustworthiness. This model suggests that, while providing traders with minimal guidance on strategies and allowing them to search for the best strategies may result in them finding effective strategies, this is not the only possible outcome. Somewhat surprisingly, effective trust strategies acted to contain unreliability, rather than dishonesty.