Speaker:

Nihar Shah, Associate Professor, Machine Learning and Computer Science Departments, Carnegie Mellon University

Title:

AI meets Peer Review: The Ace, the Cheat, the Bluff, and the Tell

Abstract:

We will discuss four facets of AI in Peer Review:
(1) The Ace: Evaluating AI in peer review, and where it surpasses human reviewers;
(2) The Cheat: How fraudsters can game vulnerabilities in the use of AI in the review process;
(3) The Bluff: Autonomous AI scientists have great promise, but also suffer from critical methodological pitfalls;
(4) The Tell: Detecting LLM generated reviews with mathematical guarantees on the family-wise error rates.

The talk will build on this blog post.

Speaker Bio:

Nihar B. Shah is an Associate Professor in the Machine Learning and Computer Science departments at CMU. His research is on the evaluation of science and the science of evaluation. His group develops computational tools with strong theoretical guarantees, and also designs and conducts controlled experiments for evidence-based policy design. His work has been used in the review of several hundred thousand papers and thousands of proposals, in over two hundred venues. He is a recipient of the 2017 David J. Sakrison memorial prize from EECS Berkeley for a "truly outstanding and innovative PhD thesis", the Young Alumnus Medal from the Indian Institute of Science, a JP Morgan faculty research award, Google Research Scholar Award, an NSF CAREER Award 2020-25, and several Best Paper Awards.

Location:

In person: Chambers Hall, 600 Foster Street, Lower Level
Remote option: https://northwestern.zoom.us/j/99003479328

About the Speaker Series:

Wednesdays@NICO is a vibrant weekly seminar series focusing broadly on the topics of complex systems, networks, and artificial intelligence. It brings together attendees ranging from graduate students to senior faculty who span all of the schools across Northwestern, from applied math to sociology to biology and every discipline in-between. Please visit: https://bit.ly/WedatNICO for information on future speakers.

Related Info

Speaker:

Michelle Birkett, Associate Professor, Department of Medical Social Sciences, Feinberg School of Medicine, Northwestern University

Title:

TBA

Abstract:

TBA

Speaker Bio:

Michelle Birkett is a tenured associate professor in the Departments of Medical Social Sciences and Preventive Medicine. She directs the Center for Computational and Social Sciences in Health (COMPASS) within the Institute for Artificial Intelligence in Medicine (I.AIM), which aims to foster connection between data science, social science and population health across Northwestern. For nearly 10 years she directed the CONNECT Complex Systems and Health Disparities Research Program within the IMPACT Institute. Dr. Birkett’s research uses network and quantitative methodologies to understand the social contextual influence of stigma on the health and wellbeing of marginalized populations, and in particular, sexual and gender minority youth. This work is influenced by a multilevel perspective of health that considers direct and indirect influences of multiple levels of the social and physical environment. This multilevel approach to understanding health underlies her interest in network data and her commitment to conducting research that leads to social change at multiple levels of society to eliminate health disparities.

Dr. Birkett has led multiple NIH-funded projects and has a wealth of expertise in the collection and analysis of network data. She is an NIH Career Awardee for her work understanding network, multilevel, and contextual influences on racial disparities in HIV within young men who have sex with men. She also directs Network Canvas, a software development project which seeks to simplify the collection and streamline the management of social data, thereby allowing health researchers to assess more nuanced associations between social contextual factors and disease. An author of more than 85 articles, Birkett's work has appeared in academic journals like The American Journal of Public Health, The Journal of Acquired Immune Deficiency Syndrome, The Journal of Adolescent Health, Social Networks, as well as popular outlets such as The Atlantic, Reuters, and Wired. In 2018 she was selected as an inaugural member of the New Voices Program of The National Academies of Sciences, Engineering, and Medicine, a initiative to promote new and diverse scientific voices within the National Academies.

Location:

In person: Chambers Hall, 600 Foster Street, Lower Level
Remote option: https://northwestern.zoom.us/j/92223012626

About the Speaker Series:

Wednesdays@NICO is a vibrant weekly seminar series focusing broadly on the topics of complex systems, networks, and artificial intelligence. It brings together attendees ranging from graduate students to senior faculty who span all of the schools across Northwestern, from applied math to sociology to biology and every discipline in-between. Please visit: https://bit.ly/WedatNICO for information on future speakers.

Related Info

Speaker:

Daniel B. Stouffer, Research Group Leader, Department of Evolutionary and Integrative Ecology, the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB), Berlin, Germany

Title:

Conceptual and Theoretical Challenges in the Study of Multi-Species Coexistence

Abstract:

The population dynamics of most ecological communities unfold on temporal scales that cannot be fully studied in the laboratory or field. The generation times of trees, for example, are so long and varied that we may need to wait decades to determine how a whole, interconnected forest community responds to a changing climate. Many researchers thus use models to generate predictions that go beyond the bounds of what is experimentally tractable. To do so, it has become common to follow the "model-paramerisation paradigm". For example, a researcher interested in forest dynamics would not conduct long-term experiments to directly probe whether one tree species is ever competitively excluded by any other(s). Instead, they would use data from shorter-term experiments to estimate the parameters of a presupposed model, and then study whether or not their empirically parameterised model predicts competitive exclusion or coexistence. As powerful as this perspective has proven to be, it routinely hinges on multiple key assumptions that limit its versatility. I will describe recent and ongoing work that challenges these assumptions, while also describing some unexpected hurdles encountered along the way.

Speaker Bio:

Daniel B. Stouffer is a Research Group Leader in the Department of Evolutionary and Integrative Ecology at the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB), Berlin, Germany. His group adopts a variety of computational, statistical, and analytical approaches to overcome ecological communities' innate complexity while exploring fundamental biological questions. They work on a variety of topics and systems and are particularly interested in understanding the emergent ecological and evolutionary consequences that arise due to interactions between species.

Location:

In person: Chambers Hall, 600 Foster Street, Lower Level
Remote option: https://northwestern.zoom.us/j/98364690035

About the Speaker Series:

Wednesdays@NICO is a vibrant weekly seminar series focusing broadly on the topics of complex systems, networks, and artificial intelligence. It brings together attendees ranging from graduate students to senior faculty who span all of the schools across Northwestern, from applied math to sociology to biology and every discipline in-between. Please visit: https://bit.ly/WedatNICO for information on future speakers.

Related Info

Speaker:

Hamsa Bastani, Associate Professor of Operations, Information, and Decisions, and Statistics and Data Science, Wharton School, University of Pennsylvania

Title:

Challenges in Achieving Human-AI Collaboration

Abstract:

TBA

Speaker Bio:

Hamsa Sridhar Bastani is an Associate Professor of Operations, Information, and Decisions (OID) and Statistics and Data Science at the Wharton School of the University of Pennsylvania, where she co-direct the Wharton Healthcare Analytics Lab. Her research sits at the intersection of machine learning, operations research, and economics. She studies how to design, deploy, and evaluate AI systems that empower human decision-makers and improve societal outcomes.

Professor Bastani aims to combine methodological depth with implementation in consequential environments. She has worked with national governments to deploy algorithms at the country scale for targeted border COVID-19 screening and essential medicine access, and has co-led one of the first large field studies of generative AI tutors in high school mathematics. She studies both the mathematical properties of algorithms and the way people respond to them.

Her research has been published in leading outlets including Nature, Management Science, Operations Research, and PNAS, and has garnered numerous recognitions, including the Wagner Prize for Excellence in Operations Research, the INFORMS Pierskalla Award for best healthcare paper, and the George Nicholson Prize. Previously, she graduated summa cum laude from Harvard in 2012 with an A.M. in physics and an A.B. in physics and mathematics, completed her PhD in Stanford's Electrical Engineering department under the supervision of Mohsen Bayati, and spent a year as a Herman Goldstine postdoctoral fellow at IBM Research.

Location:

In person: Chambers Hall, 600 Foster Street, Lower Level
Remote option: https://northwestern.zoom.us/j/99847338986

About the Speaker Series:

Wednesdays@NICO is a vibrant weekly seminar series focusing broadly on the topics of complex systems, networks, and artificial intelligence. It brings together attendees ranging from graduate students to senior faculty who span all of the schools across Northwestern, from applied math to sociology to biology and every discipline in-between. Please visit: https://bit.ly/WedatNICO for information on future speakers.

Related Info

NICO is looking for participants to share 12-15 minute lightning talks on their current research. To sign up, please fill out this short survey. These are open to Northwestern graduate students, postdoctoral fellows, and visiting scholars!

All talks will be given in person at Chambers Hall and this event will also be livestreamed via Zoom.

Speakers will be balanced based on their topics/disciplines in order to provide a broad representation of the research activities at NICO

Location:

In person: Chambers Hall, 600 Foster Street, Lower Level
Remote option: https://northwestern.zoom.us/j/98031689779

About the Speaker Series:

Wednesdays@NICO is a vibrant weekly seminar series focusing broadly on the topics of complex systems, networks, and artificial intelligence. It brings together attendees ranging from graduate students to senior faculty who span all of the schools across Northwestern, from applied math to sociology to biology and every discipline in-between. Please visit: https://bit.ly/WedatNICO for information on future speakers.

Related Info