Date: October 15, 2024 | See event page
The National Institute of Statistical Sciences (NISS) hosted the inaugural session of its new webinar series, "AI, StAtIstics, and Data Science in Practice," with an inspiring keynote by Professor Bin Yu of the University of California, Berkeley. David S. Matteson, Director of NISS, opened the event by introducing the NISS Collaboratory (CoLab) and outlining its ambitious vision to foster interdisciplinary collaboration and innovation within the statistical and data science communities.
Matteson emphasized the CoLab’s mission to strengthen partnerships, support ongoing research, and address critical, unmet needs in emerging areas. "We aim to avoid duplication by collaborating where possible but step in when leadership is needed," he explained. Among CoLab's initiatives are programs such as Statistics Serving Society (S3), focusing on timely issues like the opioid crisis and maternal mortality rates, and the NISS New Researchers Network, an expansion of the successful Graduate Student Network.
Pioneering Topics and Initiatives at NISS
Matteson highlighted several forward-thinking projects, including:
- AI and StAtIstics: A focus on bridging artificial intelligence and statistical sciences through events like AI Day and new series celebrating cross-disciplinary collaboration.
- Public Health and Environment: Addressing climate change and public health crises through data-driven decision-making.
- Visualization: Revitalizing NISS’s longstanding strength in data visualization with new graphics and design competitions launching in the spring.
He also announced an exciting upcoming series on Collaborative Data Science, which will bring statisticians, data scientists, engineers, and scientists together to explore shared challenges and solutions.
A Series to Inspire and Inform
Nancy McMillan, Data Science Research Leader within Battelle’s Health Research & Analytics Business Line, Chair of the NISS Affiliates Committee and moderator of the session, welcomed attendees to the webinar. She described the series as a platform to connect professionals, researchers, and students with leading experts in AI, statistics, and data science. Each session will feature keynote speakers addressing cutting-edge topics and practical applications, offering attendees the chance to engage directly with experts on real-world challenges and opportunities.
"We’re thrilled to kick off this series with Professor Bin Yu," McMillan said, emphasizing that the webinars will occur monthly and provide invaluable insights for participants across academia and industry.
Professor Bin Yu’s Keynote: Veridical Data Science
The spotlight of the event was Professor Bin Yu’s talk, coinciding with the release of her new book, Veridical Data Science: The Practice of Responsible Data Analysis and Decision Making. As the Chancellor’s Distinguished Professor of Statistics and Electrical Engineering and Computer Sciences at UC Berkeley, Yu shared her groundbreaking work on creating a framework for responsible and trustworthy data science.
"Veridical means truthful, and this framework aims to ensure data science results are reliable and aligned with reality," Yu explained. Developed as a response to the replication crisis, the framework incorporates the principles of predictability, computability, and stability to navigate the complexities of the data science lifecycle.
Yu’s talk delved into key concepts, including:
- The role of uncertainty quantification in data-driven decisions, particularly in critical fields like medicine.
- The importance of documenting judgment calls made throughout the data science pipeline to ensure transparency and trustworthiness.
- Practical applications of the framework through medical case studies, demonstrating its potential to address real-world challenges.
A Call to Action for the Community
Yu concluded her talk with a discussion of ongoing research and the evolution of veridical data science. "This is not a static framework—it’s philosophical, practical, and open to further development," she noted, encouraging the community to adopt and adapt the approach in diverse contexts.
The session ended with an engaging Q&A segment, where attendees posed thoughtful questions, making the webinar an interactive and enriching experience.
Looking Ahead
As the NISS Collaboratory moves forward, participants can look forward to more inspiring sessions and initiatives aimed at fostering innovation and collaboration. Stay tuned for upcoming webinars, including topics on collaborative data science, public health, and cutting-edge advancements in AI and statistics.
For more information about the NISS Collaboratory and its initiatives, visit https://www.niss.org/CoLab.
About Professor Bin Yu
Professor Bin Yu is a leading researcher whose work spans statistical machine learning, causal inference, and the integration of algorithms with domain knowledge. Her latest book, Veridical Data Science: The Practice of Responsible Data Analysis and Decision Making, is available now from MIT Press.