Data Science and AI Research Colloquia

This is an experiment, and this is new. So, we would love to hear your ideas and discuss with you any structure that would be most beneficial. As a starting point, our expectation is that each colloquium should:

• Be organized around a research theme that is broad enough to interest U-M researchers from multiple disciplines, but narrow enough to allow in-depth interaction among interested researchers. For example, 
  • “Novel NLP methods for research on misinformation on social media” is too narrow; “Novel NLP methods for social science research” is appropriate.
  • “Sequential decision making for managing chronic pain” is too narrow; “Sequential decision making in healthcare research” is appropriate.
• Happen during the fall or winter semester, lasting 1-2 days
• Include 4-6 external speakers who are prominent researchers for the theme of the colloquium, possibly with additional internal speakers
• Include seminars that benefit both faculty researchers and trainees (postdocs and graduate students), and intensive research discussions with U-M faculty
• Potentially include other components, not listed above

We will support all themes that include data science and artificial intelligence (AI) theoretical foundations, methodology and tools, and innovative application of methodologies. However, preference will be given to themes that align with the current MIDAS research “pillars”.

• Responsible Research Pillar – Enhancing Scientific and Societal Impact: Integrating methodologies and tools to enable responsible data science and AI research and enhance its scientific impact; applying data science and AI methodologies and tools to inform policy and promote social good.
• Data Pillar – Measuring and Improving Society: developing and using data science and AI methods to enable new opportunities to better understand society through new data interfaces.
• Analytics Pillar – Transforming Health Interventions: enabling the transformation of health interventions through the use of cutting edge data science and AI methods.
• Emerging Pillar – Cultivating New Strengths: research in areas that are, or are expected to be, national priorities and / or U-M strengths, and can be significantly boosted with data science and AI.
• Methodological Foundations – Advancing Data Science and AI methodologies in ways that could have downstream benefits for the above four pillars.

Three to four proposals will be chosen for the fall 2022 semester. (We expect to have a similar call for winter 2023. However, that will depend very much on our experience with the current call for fall 2022).

Colloquium organizers will be expected to:

• Take overall responsibility for organizing the event;
• Invite external and internal speakers;
• Convene U-M faculty attendees;
• Lead the design of the event program, with input from the MIDAS symposium and colloquium committee if needed;
• Ensure that the event takes place within the timeframe agreed upon with MIDAS.

MIDAS will:

• Cover the cost of the event, including travel and lodging for external speakers, venue, food for attendees and other reasonable expenses;
• Provide logistical support, including booking venue and travel, and on-site staff support;
• Advertise the event to the campus community.

• MIDAS symposium and colloquium committee will evaluate the proposals based on:
  • The scientific significance of the proposed theme;
  • The interdisciplinarity of the theme.
  • The potential of the colloquium to stimulate research discussions that will lead to innovative ideas and collaborations.
  • The fit with MIDAS pillars;
• We will then reach out to the faculty organizers to determine the exact dates of the event, the event budget, and reach an agreement on the responsibilities of the faculty organizers and of MIDAS.
• If needed, we may also discuss with the faculty organizers to tune the event theme.  We expect in most cases this will be either to broaden the theme or narrow the theme.