The 2019 Discover Innovate Impact (DII) National Data Science Challenge winning teams were recently announced and the GuanLab Team (Yuanfang Guan and Xianghao Chen) won the Task 1, Sepsis Onset Prediction prize. This was a national data science challenge established to advance human health through machine learning hosted by The University of Texas (UTHealth) School of Biomedical Informatics, sponsored by Cerner Corporation, and powered by AWS. Congratulations to GuanLab Team!
By Alex Piazza
Data science is an important tool that can help researchers tackle important societal challenges ranging from mobility and health to public safety and education.
But data science techniques and technologies also pose enormous potential for harm by reinforcing inequity and leaking private information. As a result, many sensitive datasets are restricted from research use, impeding progress in areas that impact society.
The University of Michigan, with a $2 million grant from the National Science Foundation (NSF), plans to establish a framework for a national institute that would enable research using sensitive data, while preventing misuse and misinterpretation.
“Data science has proven time and time again to be an invaluable resource when addressing emerging challenges and opportunities in areas of broad potential impact,” said H.V. Jagadish, director of the Michigan Institute for Data Science. “But having access to information comes with a great deal of responsibility, so our first priority is to ensure data science is not misused to disproportionately harm underrepresented groups.”
U-M researchers will partner with colleagues at New York University and the University of Washington over the next two years to deploy new techniques and technologies that enable responsible data science, while establishing an interdisciplinary community focused on the study, design, deployment and assessment of equitable data systems.
Equity is an important facet of data science that NSF aims to strengthen in the coming years, as the federal agency partners with universities such as U-M to enable new modes of data-driven discovery that will transform the frontiers of science and engineering.
The centerpiece of its ongoing effort, called Harnessing the Data Revolution at NSF, is the development of national institutes that address multidisciplinary problems in big data. U-M will help lay the groundwork for developing these institutes, which will eventually serve as a point of convergence for researchers from multiple disciplines to share expertise and address pressing challenges in data science.
“Information is being gathered about all of us, from our Google searches and online purchases to property tax records and social media activity,” said Margaret Levenstein, director of the Inter-university Consortium for Political and Social Research at U-M, which maintains the world’s oldest and largest archive of research and instructional data for the social and behavioral sciences. “You would assume the usage of data to be accurate and fair, but that is not always the case. That is why building a framework is so important because, in order for us to harness the enormous potential of big data, we need to ensure equity and privacy.”
H.V. Jagadish (U-M) is the principal investigator on this grant. Robert Hampshire (U-M), Bill Howe (UW), Margaret Levenstein (U-M) and Julia Stoyanovich (NYU) are co-principal investigators.
Dr. Robert Hampshire, MIDAS core faculty and Associate Professor of Public Policy at the Ford School, and his team, receives nearly $1 million in funding from the National Science Foundation’s Convergence Accelerator. The team leaders also include MIDAS faculty members Carol Flannagan, H.V. Jagadish and Margaret Levenstein. Read more at http://fordschool.umich.edu/news/2019/hampshire-receives-national-science-foundation-convergence-accelerator-grant.
MIDAS affiliated faculty and Associate Professor in Computer Science and Engineering, Dr. Mike Cafarella, receives funding from the National Science Foundation, in its program of Convergence Accelerator in Harnessing the Data Revolution. This project, “Simultaneous Knowledge Network Programming and Extraction”, is a direct result of his team’s project funded by MIDAS. Read more at https://www.nsf.gov/od/oia/convergence-accelerator/index.jsp.
See the publication at https://www.nature.com/articles/s41591-019-0548-6
For information on the press release: https://precisionhealth.umich.edu/news-events/features/taking-machine-learning-models-in-health-care-from-concept-to-bedside/
Seven outstanding young data scientists from the US, Asia and Europe will join the Michigan Institute for Data Science (MIDAS) at the University of Michigan (U-M), as the inaugural cohort of the Michigan Data Science Fellows program. They will work at the boundaries of data science methods and domain sciences in an intellectually vibrant environment and develop collaborative relationships with the U-M data science community. The Fellows and their data science application areas are:
- Arya Farahi, coming from Carnegie Mellon University: Cosmology and its intersection with fundamental physics.
- Qianying (Ruby) Lin, coming from Hong Kong Polytechnic University: Epidemic inferences and trends.
- Patrick Park, currently at U-M: Structure and evolution of large-scale human social networks.
- Elyas Sabeti, currently at U-M: Theory and algorithms for the analysis of medical Big Data.
- Maria Veiga, coming from the University of Zurich: Developing techniques for multi-scale modeling.
- Edgar Vivanco (joint postdoctoral fellow with the National Center for Institutional Diversity), coming from Stanford University: Utilizing machine learning to examine how colonial-era institutions and contemporary criminal violence shape economic under-performance.
- Blair Winograd, currently at U-M: working with M-Write to combine conceptual writing prompts, automated peer review, natural language processing, and automated personalized feedback to create an infrastructure for writing at scale.
The two-year Fellows Program accepts recent PhDs who are stars in their respective fields and whose work is in data science. They are expected to be more independent than the average postdoctoral researchers at the same career juncture; however, each Fellow also has two faculty sponsors, one a methodology expert, and the other an expert in an application domain, to ensure scientific and career guidance.
The Fellows program is a new component of MIDAS’ effort to catalyze the transformative use of Data Science in a wide range of disciplines to achieve lasting societal impact, through research, education, outreach and partnership. “This is the first postdoctoral training program at U-M, and one of the few in the nation, with data science as the explicit focus,” says Dr. H.V. Jagadish, MIDAS Director and Professor of Computer Science and Engineering, “and we hope this program will foster the next generation of data science leaders with both a strong scientific vision and a commitment to using data science for positive societal impact.”
One of the Fellows, Elyas Sabeti, expressed great enthusiasm: “This is such a unique opportunity. It’s amazing that I will be working side by side with people who study Physics, Education, Political Science… I can’t wait to find out how many great ideas we can come up with together.”
For more information on the Fellows, please click here.
The National Science Foundation recently announced a second phase of funding for its regional Big Data Innovation Hub program, which comprises a growing network of partners investing in data and data sciences to address grand challenges for society and science.
As part of a four year, $4 million award, the University of Michigan will collaborate with Indiana University, Iowa State University, the University of Minnesota – Twin Cities and the University of North Dakota to establish priority focus areas for the Midwest Big Data Hub (MBDH). The National Center for Supercomputing Applications at the University of Illinois, Urbana-Champaign will continue to lead the hub.
“The Midwest Big Data Hub community engages dozens of regional and national universities, transdisciplinary scholars and industry partners in tackling complex data-driven challenges, developing unique data science courses and training resources, as well as promoting data-science-for-social-good,” said H. V. Jagadish, director of the Michigan Institute for Data Science (MIDAS) and the Bernard A. Galler Collegiate Professor of Electrical Engineering and Computer Science at U-M.
U-M plays an important role in coordinating efforts across campuses so that researchers can harness the power of big data to address important issues related to:
- Advanced materials manufacturing
- Water quality
- Big data in health
- Digital agriculture
- Smart, connected and resilient communities.
“The University of Michigan is leading in biomedical and health research, as well as development and educational activities, while actively participating in other priority areas like water quality and smart manufacturing,” said Ivo Dinov, U-M professor of computational medicine and bioinformatics, who also serves as the associate director for education at MIDAS. “The Michigan data science and predictive health analytics research and development will focus on developing advanced clinical decision support systems that can be used to diagnose, track and predict the onset of devastating disorders like cognitive or memory decline, mental health and cancer.”
The Midwest hub, which launched in 2015 with support from NSF, serves a 12-state region that comprises Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota and Wisconsin. The hub leads cross-cutting initiatives for broadening participation in data science education, cyberinfrastructure for research data management and cybersecurity issues around big data. By leading initiatives in data science education and workforce development, the hub aims to increase data science capacity within the region. The NSF also funds regional hubs in the Northeast, South and West, which together cover the entire United States.
“Developing innovative, effective solutions to grand challenges requires linking scientists and engineers with local communities,” said Jim Kurose, NSF assistant director for computer and information science and engineering. “The Big Data Hubs provide the glue to achieve those links, bringing together teams of data science researchers with cities, municipalities and anchor institutions.”
Learn more about the Midwest Big Data Hub at midwestbigdatahub.org.
Nia Dowell, who was a postdoc fully funded by the MIDAS holistic modeling of education project working with MIDAS faculty Christopher Brooks and Tim McKay, was the the lead author on the Best Paper representing Education for All at the annual Artificial Intelligence in Educational (AIED2019) conference. This conference is one of the oldest in the field, going back to the late 80’s. Her work is focused on analytics of education and inclusion. The title of the paper is “Promoting Inclusivity Through Time-Dynamic Discourse Analysis in Digitally-Mediated Collaborative Learning”. Read more about the paper.
Dr. Sue Hammoud and the Michigan Center for Single-Cell Genomic Data Analytics team received grants from the Open Philanthropy Project ($2.5 million) and Chan Zuckerberg Initiative ($1.2 million).
The Open Philanthropy Project awarded a grant of $2,500,000 over four years to the University of Michigan to support research by Drs. Sue Hammoud and Jun Li on mammalian gamete development in March of 2019. The research would be specifically focused on development of gametes from stem cells.
Progress in this area could eventually enable people with fertility challenges to have children and could eventually help reduce the incidence of a wide variety of high-burden disorders (such as heart disease, chronic pain, depression, and Alzheimer’s disease) and promote other positive outcomes. Dr. Hammoud’s research is amongst the most promising our science team has encountered so far in this field.
The Chan Zuckerberg Initiative (CZI) awarded $1.2 million to Drs. Sue Hammoud, Jun Li, Erica Marsh and Ariella Shikanov at the University of Michigan. This project will establish a human cell atlas of the female reproductive system, focusing on the ovaries, fallopian tube, and uterus.