REU Site: The future of discovery: training students to build and apply open source machine learning models and tools

Machine learning is a powerful tool that has been successfully applied to a variety of problems that until recently were deemed too difficult or impossible for computers to solve. This REU Site project gives participating students experience in many aspects of machine learning, ranging from developing open source machine learning models and tools to applying them in the real world.

The work carried out by the students will lead to research advances in the fields of these projects and the models and tools they develop will be open-source, leading to them being available to other fields where these models can be used to make additional advances. Machine learning is an emerging field with limitless opportunities to design innovative services and products that will enhance the lives of billions of people, help to address emerging challenges in climate, food, water, energy, transportation, and healthcare, and advance science and engineering discoveries in ways unimaginable today.

The project contributes to the development of a highly specialized workforce trained to utilize advanced machine learning methods, and to contribute to open source software. Students from diverse backgrounds and computational/data-oriented disciplines are being trained to apply machine learning and to participate in research where these tools are at the center of scientific discovery, preparing them to apply machine learning methods in other fields and providing them with the foundation and motivation to pursue advanced graduate studies.

This project serves NSF's mission by promoting the progress of science and advancing national health, prosperity and welfare.

The goals of this project are to train undergraduate students, focusing on those from minority serving institutions, in machine learning and open source software, where they will then apply these skills to mentor-guided research projects. This is an on-site summer program at the University of Illinois that brings to campus 10 students per year and is based on matching their preferences and interests to those of a group of mentors, so that each student works with a pair of mentors, one from the project's research area and the other with expertise in machine learning.

This program increases the students' knowledge of research and graduate school, and in many cases, stimulates their interest in continuing to graduate school, while in other cases, trains students with skills that enable them to seek data science and data analysis jobs in industry, increasing diversity in these graduate programs and in industry. By their presence in the program as continuing undergraduates, when the students return to their university, they will build a relationship between Illinois and that university, their faculty, and their peers that encourages future students to participate in the program and provides the basis for future joint research projects.

This 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.