New Traineeship will Train Researchers at the Intersection of Astrophysics and Data Science

School of Physics and Astronomy Professor Vuk Mandic is the principal investigator of a National Science Foundation Research Traineeship designed to train graduate students in data science, in the context of the nascent field of Multi-Messenger Astrophysics. The new training program will be conducted by an interdisciplinary group of faculty, cutting across two Colleges and five programs and will provide a total of 30 annual stipends for graduate students, during 2020-2024, each in the amount of $34,000 plus tuition and fees.

Faculty members involved are Dan Boley (Computer Science), Lucy Fortson (Physics), Jarvis Haupt (Electrical Engineering), Galin Jones (Statistics), Pat Kelly (MIfA), Vipin Kumar (Computer Science), Vuk Mandic (Physics), Claudia Scarlata (MIfA), and Xiaotong Shen (Statistics).

On August 17, 2017, a merger of two neutron stars was simultaneously observed as a gravitational wave signal by LIGO and Virgo detectors, and as a Gamma Ray Burst by Fermi and INTEGRAL satellites. The event was then followed up in all parts of the electromagnetic spectrum by numerous terrestrial and space-borne detectors and telescopes. This event is often regarded as the birth of the Multi-Messenger Astrophysics, where astrophysical events are observed by multiple messengers (gravitational waves, light, neutrinos, cosmic rays), enabling previously impossible scientific inquiries. The 2017 event has demonstrated this science potential by enabling novel tests of gravity, measurements of the expansion of the universe, probes of the state of matter in neutron stars, and understanding of the origin of the heaviest elements in the periodic table.

In the coming years, vast amounts of new data are expected from existing and upcoming gravitational wave detectors, telescopes, gamma ray detectors, and neutrino detectors. The size, complexity, and diversity of the new datasets will be unlike anything seen in Astrophysics. The National Science Foundation Research Traineeship awarded to the University of Minnesota will facilitate the processing and analysis of these data by training MS and PhD students in modern data science techniques. The Traineeship will use the nascent field of Multi-Messenger Astrophysics as a training ground to prepare students for the challenges of the modern data-driven workforce in both industry and academia.

The trainees will pursue unique research opportunities by working in interdisciplinary trainee teams, mentored by interdisciplinary faculty mentoring teams. They will use the most modern data analytics techniques, such as machine learning, deep learning, Bayesian inference and others, to address some of the pressing questions in Multi-Messenger Astrophysics. Their training will include a series of workshops, seminars, symposia, journal club, outreach activities, and Fall retreats, designed to provide the trainees with a broader exposure to data science techniques, tools, and applications, and with opportunities for development of professional communications skills, leadership and public skills, and career development, vision, and impact skills. The trainees will pursue summer research and industry internships with local and national companies and research organizations. A series of activities will be pursued to ensure full participation of underrepresented minorities and women, and to develop a strong community around this program that will provide the support structure for the trainees, help them move through the program efficiently, and provide them with career opportunities post-graduation. The best practices developed by this program will be publicly available for adoption by other STEM graduate programs across the nation.