Electrical Engineering and Computer Science

The Future of Data Science: Kicking Off U-M’s Proactive Step into an Exploding Field

Researchers from around the nation gathered at Rackham on October 6 to celebrate the official launch of Michigan’s $100M Data Science Initiative. Central to this program is the new Michigan Institute for Data Science (MIDAS), which aims to make sense of the massive datasets researchers in every field now have at their disposal.

The symposium, titled “The Future of Data Science: A Convergence of Academia, Industry, and Government,” was an all-day event featuring representatives of many major industries and academic institutions. Prof. Eric Michielssen, Associate Vice President of Research Computing and Director of the Michigan Institute for Computational Discovery and Engineering, summarized the broad goals of the UM Data Science Initiative.  MIDAS Co-Directors Al Hero, R. Jamison and Betty Williams Professor of Engineering, and  Brian Athey, Michael A. Savageau Collegiate Professor, offered an overview of the institute and its goals.

MIDAS aims to support ongoing research projects involving data science across the University, as well as to facilitate training for a new generation of data science researchers. It will focus its efforts on four key areas – learning analytics, transportation, computational social science, and health sciences – but projects could emerge in any field.


Al Hero, Co-Director of MIDAS. Hero is the  R. Jamison and Betty Williams Professor of Engineering and a professor of Electrical Engineering and Computer Science

“This initiative deals with the tremendous firehose of data coming out of Michigan research, the amount of which is now increasing exponentially,” Prof. Hero said at the event. To meet that demand, MIDAS embraces interdisciplinary collaboration. Currently, about 150 faculty across the U-M campuses are affiliated with the institute.

To help meet the demand for a trained workforce in data science, the institute offers a graduate certificate in data science, and is developing a 5-year combined Bachelor’s/Master’s program. The institute also has plans to launch an online version of the graduate certificate program as well as four Massive Online Open Courses (MOOCs) in various data science areas.

In addition, MIDAS taps into a thorough network of industry partners, described by Ivo Dinov, MIDAS’ Associate Director of Education and Training, as the “MIDAS Medusa.” These partners will provide support on research projects and in some cases, contribute their own massive datasets as test cases.


The "MIDAS Medusa" of industry partnerships.

Other speakers highlighted the breadth of data science research already ongoing in industry and academia, including the social sciences, healthcare, astronomy, the automotive industry, app development, and more.

Four sets of themed talks and panel sessions focused on the different aspects of the Data Science Initiative: 1) advancing data science methodologies, 2) data science education and training, 3) data science challenges and opportunities, and 4) data science community partnerships.


(From left) Al Hero, Susan Murphy, Kathleen McKeown, and Robert Nowak participate in the day's first panel discussion, on data science methodologies.

This initiative comes at a time of heightened interest in computing and data sciences. Prof. Michielssen compared U-M’s mission to President Obama’s recently announced National Strategic Computing Initiative, which seeks to apply exaFLOPS [one billion billion operations per second] of computing power to exabytes of data. “Data and computational sciences are just two sides of the same coin,” says Prof. Michielssen, “Big Data research and high-performance computing work together to push the base of discovery to new levels.”

With the MIDAS touch, U-M promises to play a big part in the future of data science. As Prof. Michielssen put it, “MIDAS will be the Big House of Big Data.”


MIDAS Symposium Speakers (in order of first appearance):

Eric Michielssen, Associate Vice President for Advanced Research Computing
Martha Pollack, Provost
S. Jack Hu, Interim Vice President for Research
Brian Athey, Co-Director of MIDAS and Michael A. Savageau Collegiate Professor
Al Hero, Co-Director of MIDAS and R. Jamison and Betty Williams Professor of Engineering
Robert Nowak, McFarland-Bascom Professor in Engineering, University of Wisconsin-Madison
Susan Murphy, H.E. Robbins Distinguished University Professor of Statistics, U-M
Kathleen McKeown, Director, Institute for Data Sciences and Engineering, Columbia University
Ivo Dinov, Associate Director for Education and Training, MIDAS; Associate Professor of Human Behavior and Biological Sciences; Director, Statistics Online Computational Resource
Erin Shellman, Research Scientists, Amazon Web Services
Patrick Harrington, Co-Founder and Chief Data Scientist, Comp Genome
Nandit Soparkar, Chief Executive Officer, Ubiquiti
Daniel L. Goroff, VP and Program Director, Alfred P. Sloan Foundation
George Poste, Regents’ Professor and Del E. Webb Chair in Health Innovation, Arizona State University
Bror Saxberg, Chief Learning Officer, Kaplan Learning and Innovation
Kathleen M. Carley, Professor of Computation, Organization, and Society, Carnegie Mellon University
Jonathan H. Owen, Director of Operations Research, VP of Practice, INFORMS, General Motors
Edward Seidel, Founder Professor, Departments of Physics and Astronomy, University of Illinois at Urbana-Champaign
Ratna Chinnam, Professor of Engineering, Wayne State University
Keith Elliston, Chief Executive Officer, tranSMART Foundation


Posted: October 28, 2015

More Info

Videos
The Future of Data Science: Watch every talk from the event in high quality video, download slides, read abstracts of the talks

In the News

MIDAS announces competition for data science project funding, The University Record, October 15, 2015

Michigan Institute for Data Science: Bringing the MIDAS Touch to Big Data, EECS Department,.September 8, 2015.

U-M launching $100 million Data Science Initiative, The University Record, September 8, 2015.