REU Site: CAAR: Combinatorics, Algorithms, and AI Applied to Real Problems

REU 网站：CAAR：组合学、算法和人工智能应用于实际问题

• 批准号: 2150382
• 负责人: William Gasarch
• 金额: $ 42.21万
• 依托单位: University of Maryland, College Park
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-03-01 至 2025-02-28
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2150382&HistoricalAwards=false
• 关键词: REU; Site; CAAR; Combinatorics; Algorithms

This renewal REU Site program will recruit 10 undergraduate students to come to the University of Maryland at College Park where they will work for 10 weeks on team research projects with mentors. The students will be offered a variety of research projects and will also be given a strong voice in the creation of their own projects. Several will develop new techniques and analyses of machine learning (ML) systems, which use probability and statistics to find patterns in data. Others will use cryptography to build secure systems. Research projects in following topical areas will be offered: Cryptography and Security; Fair Allocation; Trustworthy Machine Learning; and Neural Networks and Finance. These projects broadly focus on combining theory and practice and students attending the program will leave it with the ability to look at a real world problem and apply theoretical techniques to solving it. There will be a special effort to recruit students in the program from non-research-focused schools and from traditionally underrepresented groups. The PIs plan several recruiting trips (in person or on zoom) to such schools to attract such students. This program will give many of the students a chance to do research, which they otherwise would not have. This program will prepare all of the students for graduate school. Social activities include trips, hikes, panel discussions about professional development, and other planned interactions with other REU students, current graduate students, and facultyThis award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

这个更新REU网站计划将招募10名本科生来到大学公园的马里兰州大学，在那里他们将与导师一起进行为期10周的团队研究项目。学生将获得各种研究项目，并将在创建自己的项目中获得强有力的发言权。一些人将开发机器学习（ML）系统的新技术和分析，这些系统使用概率和统计来发现数据中的模式。其他人将使用密码学来构建安全系统。将提供以下主题领域的研究项目：密码学和安全;公平分配;值得信赖的机器学习;神经网络和金融。这些项目主要侧重于理论与实践的结合，参加该项目的学生将具备研究真实的世界问题并应用理论技术解决问题的能力。该项目将特别努力从非研究型学校和传统上代表性不足的群体中招收学生。PI计划几次招聘旅行（亲自或放大）到这些学校，以吸引这些学生。 这个项目将给许多学生一个做研究的机会，否则他们就不会有。该计划将为所有学生准备研究生院。社会活动包括旅行，徒步旅行，小组讨论有关专业发展，以及其他计划与其他REU学生，目前的研究生和facultyThis奖反映了NSF的法定使命的互动，并已被认为是值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估的支持。

项目成果

Robustness Disparities in Face Detection

• DOI: 10.48550/arxiv.2211.15937
• 发表时间: 2022-11
• 期刊: ArXiv
• 作者: Samuel Dooley;George Z. Wei;T. Goldstein;John P. Dickerson

Learning Revenue-Maximizing Auctions With Differentiable Matching

• 发表时间: 2021-06
• 期刊: ArXiv
• 作者: Michael J. Curry;Uro Lyi;T. Goldstein;John P. Dickerson

Cliff Diving: Exploring Reward Surfaces in Reinforcement Learning Environments

• DOI: 10.48550/arxiv.2205.07015
• 发表时间: 2022-05
• 期刊: ArXiv
• 作者: Ryan Sullivan;J. K. Terry;Benjamin Black;John P. Dickerson

On the Generalizability and Predictability of Recommender Systems

• DOI: 10.48550/arxiv.2206.11886
• 发表时间: 2022-06
• 期刊: ArXiv
• 作者: Duncan C. McElfresh;Sujay Khandagale;Jonathan Valverde;John P. Dickerson;Colin White

The Dichotomous Affiliate Stable Matching Problem: Approval-Based Matching with Applicant-Employer Relations

• DOI: 10.24963/ijcai.2022/51
• 发表时间: 2022-02
• 期刊: ArXiv
• 作者: Marina Knittel;Samuel Dooley;John P. Dickerson