REU Site: Multi-Agent Simulations of Social Systems

REU 站点：社会系统的多智能体模拟

• 批准号: 0755405
• 负责人: Yu Zhang
• 金额: $ 20.5万
• 依托单位: Trinity University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2008
• 资助国家: 美国
• 起止时间: 2008-05-01 至 2012-04-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0755405&HistoricalAwards=false
• 关键词: REU; Site; Multi; Agent; Simulations

Abstract:This project focuses on the study of multi-agent systems (MAS) in social systems. The proposed research breaks into three parallelprojects: 1) multi-agent systems that support natural human decision- making, 2) a simulation of urban areas and their inhabitants that can be used in the MAS, and 3) a data analysis and visualization tool for the system. Students will receive training in a variety of areas, including multi-agent systems¡¯design and implementation, spatial modeling, empirical methods for system evaluation and statistical techniques for data analysis. In addition, students will attend professional meetings, learn to assimilate the latest research from reading and discussing recent papers, and learn how to write up and present the results of their own research. This REU program will focus on strong mentoring and individual contact with the students and will include weekly sessions to help them pursue graduate studies.Intellectual Merit. The proposed research will develop a new multi- agent framework. Contributions to the MAS community will include a multi-agent system that represents dynamic interactions between agents and social structures that shape the agent¡¯s behaviors, and a decision-making mechanism that captures the common human decision- making attitudes through modeling the domain of human intuition. We will also use this MAS to simulate urban areas and their inhabitants and develop a data analysis and visualization tool for the system. Students will have the opportunity to work in depth in the core area of MAS or with the application of the core area to different disciplines in computer science.Broader Impact. This project creates a model program for liberal arts institutions that addresses the exposure to research by undergraduates as well as awareness of the pathway to graduate school. While the recruiting effort for students will be nationwide, we will emphasize the recruitment of students from underrepresented groups. Of particular interest to us are Hispanic students enrolled at the many Hispanic serving colleges and universities in South Texas. Training a more diverse engineering and science professoriate will benefit society significantly, not only in South Texas but nationally, in promoting a more equitable participation of minorities and women in the computer science disciplines.

翻译后摘要：该项目的重点是多智能体系统（MAS）在社会系统的研究。建议的研究分为三个并行项目：1）多智能体系统，支持自然的人类决策，2）模拟城市地区和他们的居民，可以在MAS中使用，和3）数据分析和可视化工具的系统。学生将接受多领域的培训，包括多智能体系统的设计和实现、空间建模、系统评估的经验方法和数据分析的统计技术。此外，学生将参加专业会议，学习从阅读和讨论最近的论文中吸收最新的研究，并学习如何撰写和展示自己的研究成果。这个REU计划将侧重于强有力的指导和与学生的个人接触，并将包括每周会议，以帮助他们追求研究生学习。本研究将发展一个新的多智能体架构.对MAS社区的贡献将包括一个多智能体系统，该系统代表智能体与塑造智能体行为的社会结构之间的动态交互，以及一个决策机制，该机制通过对人类直觉领域进行建模来捕捉人类共同的决策态度。我们还将使用该MAS来模拟城市地区及其居民，并为系统开发数据分析和可视化工具。 学生将有机会深入研究MAS的核心领域，或将核心领域应用于计算机科学的不同学科。该项目为文科院校创建了一个示范项目，旨在解决本科生对研究的接触以及对进入研究生院途径的认识。虽然招生工作将在全国范围内进行，但我们将强调从代表性不足的群体中招收学生。我们特别感兴趣的是在南德克萨斯州的许多西班牙裔服务学院和大学就读的西班牙裔学生。培养更多样化的工程和科学教授将使社会受益匪浅，不仅在南德克萨斯州，而且在全国范围内，促进少数民族和妇女更公平地参与计算机科学学科。