U-M Annual Data Science & AI Summit 2023

November 13-14, 2023

Rackham Building

915 E. Washington St., Ann Arbor

About

The U-M Data Science and AI Summit is the largest annual data science and AI event on campus. This event brings together the U-M data science and AI research community and their external collaborators to build research vision and collaboration. It also showcases the breadth and depth of U-M data science and AI research, from theory and methodology development to the transformative use of data and AI to address scientific and societal challenges in all domains. The event is free for all attendees (U-M faculty, staff, and trainees, as well as industry, government and community members).

Summit Program Speakers and Panelists Keynote Information Research Talks PODS Showcase Posters Session Program Committee

Note: all sessions held in Amphitheatre (4th) floor, with overflow seating in the East and West Conference Rooms, unless otherwise noted below. For facility and accessibility information, click here.

8:30 am

Registration and check-in (ground floor lobby): check in early! The first 150 registrants to check in will receive special MIDAS swag.

8:45 am

Opening Remarks

Rebecca Cunningham, Vice President Emerita for Research; William G. Barsan Collegiate Professor of Emergency Medicine, University of Michigan
H. V. Jagadish, Director, Michigan Institute for Data Science; Edgar F Codd Distinguished University Professor and Bernard A Galler Collegiate Professor of Electrical Engineering and Computer Science, University of Michigan

View recording

9:00 am

Keynote – Suresh Venkatasubramanian, Responsible AI: from a conference workshop to the White House

Director, Center for Technological Responsibility, Reimagination, and Redesign, Data Science Institute, Brown University; Former Asst. Dir for Science and Justice, White House Office of Science and Technology Policy

The first FATML workshop was hosted at NeurIPS in Dec 2014, in a conference room that held 20 people. On Oct 30, 2023, in the East Room of the White House, the President signed an executive order on safe, secure, and trustworthy AI in front of over 100 people, with thousands more watching from the outside.

I was there for both of these events. And for many events in between. In this talk, I’ll talk about the journey from algorithmic fairness to AI governance, the coalition that advocated so fiercely for protections for people, and what comes next in the world of responsible AI.

Session Chair: Karandeep Singh (Associate Professor of Learning Health Sciences, Internal Medicine, Urology, and Information; Associate Chief Medical Information Officer of Artificial Intelligence, Medical School)

View recording

10:00 am

Propelling Original Data Science Grant 2023 Awardees Showcase

View recording

11:10 am

Research Vision Talks Session 1 - Data Science / AI and societal impact

Benjamin Goldstein, Data for Urban Sustainability – The Good, The Bad, and The Ugly

Assistant Professor of Environment and Sustainability, School for Environment and Sustainability

Tian An Wong, Proactive Policing as Reinforcement Learning

Assistant Professor of Mathematics, Department of Mathematics and Statistics, College of Arts, Sciences and Letters, The University of Michigan-Dearborn

Donglin Zeng, Personalized Medicine Discovery through Machine Learning

Professor of Biostatistics, School of Public Health

Session Chair: Kai Zhu, Associate Professor of Environment and Sustainability, School for Environment and Sustainability and Associate Professor of Ecology and Evolutionary Biology, College of Literature, Science, and the Arts

More info

View recording

12:10 pm

Lunch, Networking, Poster Session and Poster Lightning Talks (Assembly Hall, East & West Conference Rooms)

Showcasing a wide range of data science and AI research projects and grassroots data science and AI organizations

1:45 pm

Keynote – Julianne Dalcanton, Making Sense of a Complicated Universe

Director, Center for Computational Astrophysics, Flatiron Institute

As a field of inquiry, astrophysics is both fascinating and daunting. Decades of research has created an incredibly rich legacy of both observations and theory, which has shaped our understanding of the complex Universe of stars and galaxies. However, this same research has driven home the profound interconnectedness of the underlying astrophyics, where processes that happen on one scale can shape observable features on scales ten orders of magnitude larger. In this talk, I’ll give a broad overview of this interconnected universe, and some of the ways that advances in machine learning are helping us grapple with fundamental difficulties in the field.

Session Chair: Lia Corrales (Assistant Professor of Astronomy, College of Literature, Science, and the Arts)

2:45 pm

Research Vision Talks Session 2 - methodology and the physical sciences

Christopher Miller, Using Galaxy Shapes to Develop a Debiasing Framework for Deep Learning

Professor of Astronomy and Professor of Physics, College of Literature, Science, and the Arts

Andrew Owens, Multimodal Learning from the Bottom Up

Assistant Professor of Electrical and Computer Engineering, College of Engineering

Venkat Viswanathan, Differentiable Physics

Associate Professor of Aerospace Engineering, College of Engineering

Session Chair: Elle O’Brien, Lecturer III in Information and Research Investigator, School of Information

More info

View recording

4:00 pm

University of Michigan Data Science and AI Organizations Showcase

Presentations

Christina Certo, Michigan Artificial Intelligence Laboratory (AI Lab)

Program Manager of Strategic Initiatives in AI

Henrike Florusbosch, E-Health and Artificial Intelligence (e-HAIL)

Program Manager, e-HAIL

Patrick Schloss, The Carpentries at Michigan

Program Director of Microbiology and Immunology AP&A, Frederick G Novy Collegiate Professor of Microbiome Research and Professor of Microbiology and Immunology

Vancho Kocevski, Michigan Institute for Computational Discovery and Engineering (MICDE)

Managing Director, MICDE

Daniel Forger, Michigan Center for Applied and Interdisciplinary Mathematics (MCAIM)

Director, MCAIM; Robert W. and Lynn H. Browne Professor in Science, Professor of Mathematics and Research, and Professor of Computational Medicine and Bioinformatics

Carol Janney, Consulting for Statistics, Computing and Analytics Research (CSCAR)

Data Science / CSCAR Consultant

Kelsey Ebbs, Bold Challenges

Assistant Director, Bold Challenges

Patricia Garcia, Center for Ethics, Society, and Computing (ESC)

Assistant Professor of Information, School of Information and Assistant Professor of Digital Studies Institute, College of Literature, Science, and the Arts

Ellie Abrons, Digital Studies Institute

Associate Professor of Architecture, A Alfred Taubman College of Architecture and Urban Planning; Director, Digital Studies Institute and Associate Professor in the Digital Studies Institute, College of Literature, Science, and the Arts

Julie Berson Grand, Science, Technology and Public Policy Program (STPP)

Education Program Manager, STPP, Ford School of Public Policy

Sebastian Zöllner, University of Michigan Precision Health

Co-Director, Precision Health; John G Searle Associate Professor of Biostatistics, Professor of Biostatistics, School of Public Health and Professor of Psychiatry, Medical School

Session Chair: Jing Liu (Executive Director, MIDAS)

View recording

5:30 pm

Reception and Networking (Assembly Hall)

All attendees welcome*. Come and talk with U-M data science and AI organizations, researchers, and students, as well as industry and public-sector partners.

*Note: we are not able to permit undergraduate students to attend the reception.

8:45 am

Keynote – Emre Kiciman, A New Frontier at the Intersection of Causality and LLMs

Senior Principal Researcher, Microsoft Research

Correct causal reasoning requires domain knowledge beyond observed data. Consequently, the first step to correctly frame and answer cause-and-effect questions in medicine, science, law, and engineering requires working closely with domain experts and capturing their (human) understanding of system dynamics and mechanisms. This is a labor-intensive practice, limited by expert availability, and a significant bottleneck to widespread application of causal methods.

In this talk, we will delve into the causal capabilities of large language models (LLMs), discussing recent studies and benchmarks of their ability to retrieve and apply causal knowledge, as well as the limitations of their causal reasoning capabilities. Most notably, LLMs present the first instance of general-purpose assistance for constructing causal arguments, including generating causal graphical models and identifying contextual information from natural language. This promises to reduce the necessary human effort and error in end-to-end causal inference and reasoning, broadening their practical usage. Ultimately, by capturing common sense and domain knowledge, we believe LLMs are a catalyst for a new frontier facilitating translation between real world scenarios and causal questions, and formal and data-driven methods to answer them.

Session Chair: Walter Dempsey (Assistant Professor of Biostatistics, School of Public Health and Research Assistant Professor, Survey Research Center, Institute for Social Research)

View recording

9:55 am

Research Vision Talks Session 3 - Data Science / AI for human health

Andrew Krumm, How is Surgical Training like a Computer Adaptive Test?

Assistant Professor of Learning Health Sciences, Assistant Professor of Surgery, Medical School and Assistant Professor of Information, School of Information

Elle O’Brien, Barriers to Data Science Adoption in Scientific Research

Lecturer in Information and Research Investigator, School of Information

Michele Peruzzi, One science fits most: how statistical models can uncover the ecology of cancer and the pathology of climate change

Assistant Professor of Biostatistics, School of Public Health

Chandra Sripada, Discovering Markers and Mechanisms of Mental Disorders with Natural Language Processing and Daily Thought Sampling

Theophile Raphael Research Professor of Clinical Neurosciences, Professor of Psychiatry, Medical School, Professor of Philosophy and Adjunct Professor of Psychology, College of Literature, Science, and the Arts

Session Chair: Alex Gorodetsky, Assistant Professor of Aerospace Engineering, College of Engineering

More info

View recording

11:20 am

Panel Discussion - Federal Priorities in Data Science and AI and Academic Research Opportunities

Panelists

Laura Biven

Data Science Technical Lead, Office of Data Science Strategy, National Institutes of Health

Andrew Carney

Program Manager, Advanced Research Projects Agency for Health (ARPA-H)

Michael Molnar

Director, Advanced Manufacturing National Program Office, National Institute of Standards and Technology

Hector Muñoz-Avila

Program Director and Cluster Lead, the Information Integration and Informatics Program, National Science Foundation

Alvaro Velasquez

Program Manager, Information Innovation Office, Defense Advanced Research Projects Agency

Moderator

H. V. Jagadish

Director, Michigan Institute for Data Science; Edgar F Codd Distinguished University Professor of Electrical Engineering and Computer Science; Bernard A Galler Collegiate Professor of Electrical Engineering and Computer Science

12:45 pm

Poster Awards and Closing Remarks

12:50 pm

Light Lunch and Networking (Assembly Hall & West Conference Room)

All attendees welcome

12:50 pm

Special Session: From Guidelines to Outcomes – a brainstorming session on implementing AI for your organization (East Conference Room)

Overview: While AI brings unprecedented potential to transform work, life and society, a major challenge facing organizations in the public sector, industry and academia is how to turn such potential into reality. As we explore the many uses of AI and the ethical AI principles, what technical capabilities, tools and processes are essential for an organization to implement AI in ethical ways and align with its mission to achieve optimal outcomes? In this brainstorming session co-hosted by MIDAS and Rocket Companies, we invite leaders and technical personnel from industry, government, non-profit organizations and academia to discuss roadblocks and seek solutions together.