REU Site in Computer Systems

Harvey Mudd College, Claremont CA

Investigators

Abstract

This funding renews a Research Experience for Undergraduate (REU) site at Harvey Mudd College. The site brings 10 undergraduates in order to engage them in research and encourage graduate study in computer science. The intellectual focus of the site is research in computer systems. The program provides a microcosm of the graduate experience through four broadly-scoped systems projects. Students have a chance to gain experience in all aspects of the research process, including its social aspects. The site focuses on recruiting a sufficiently diverse pool of students, including women and underrepresented minorities. The intellectual merit of the proposal rests with the leadership of an experienced team of mentors with excellent expertise and experience in the research areas. The site focuses on four main projects. The first project develops algorithms for estimating cophylogenetic trees based on biological host/parasite data. Students' work is encapsulated in a software package named Jane, currently in use by dozens of biology research groups. The second effort investigates approaches for automatic synthesis of music scores, furthering our understanding of computational music research. It leverages the Impro-Visor software, which has many thousands of users. The third project explores program analyses to support new, accessible domain-specific languages, secure information flow in industry-standard languages such as Javascript, and visualization tools for both sides of that programming spectrum. The fourth project tests machine-learning and computer vision-based algorithms that improve the state-of-the-art of performance of low-cost robots, such as sub-$300 quadcopters and laser-guided wheeled platforms. The broader impacts include the site's emphasis on the broad human and research impacts of pursuing computer science at the graduate level -- especially for participants who had not previously considered graduate work. Challenging research problems prompt students' guided development of research skills: investigation, presentation, and publication. One-on-one introductions by advisors transition to student-led talks and culminate with publications and conference experiences during or after the summer. The program's most important impact is its cultivation of the next generation of creative and enthusiastic computer science researchers. The evaluation plan and the student research artifact archiving plan should contribute to educational research on effective structures for involving undergraduate students in research.