STAtistical Methods for the Physical Sciences Research Center Launch - Keynote Talks

— 6:00pm

Location:
In Person and Virtual - ET - Simmons Auditorium B, Tepper Building and Remote Access

Speaker:
KYLE CRANMER and AMY BRAVERMAN

Two Talks

►  Dr. KYLE CRANMER      

 —  The intersection of statistics, machine learning, and the physical sciences All models are wrong, but some are useful” is a famous quip by George Box recognizing that statisticians commonly employ models that are simplifications of the real world. Similarly, physicists are notorious for invoking ‘spherical cows’ that capture the essence of a system that is conducive to reasoning or pencil-and-paper calculations. While this approach has its place, the cutting-edge of the physical sciences are characterized by highly detailed simulations with many components interacting through some underlying mechanistic model. Moreover, physical scientists are often most interested in inferring aspects of the mechanistic model itself, but this involves very challenging inverse problems. In contrast, the field of machine learning generally eschews interpretable, generative models in favor of black-box models and an optimization perspective.  I will describe how machine learning techniques, despite their black-box nature, are empowering a revolution in principled statistical inference for the physical sciences. 

Kyle Cranmer is the David R. Anderson Director of the UW-Madison Data Science Institute and a Professor of Physics with courtesy appointments in Statistics and Computer Science. He is also the Editor in Chief of the journal Machine Learning Science and Technology. Cranmer was a Professor of Physics and Data Science at NYU from 2007 – 2022. He obtained his Ph.D. in Physics from the University of Wisconsin-Madison in 2005. He was awarded the Presidential Early Career Award for Science and Engineering in 2007, the National Science Foundation's Career Award in 2009, and became a Fellow of the American Physical Society in 2021 for his work at the Large Hadron Collider. Professor Cranmer developed a framework that enables collaborative statistical modeling, which was used extensively for the discovery of the Higgs boson in 2012. His current interests are at the intersection of physics, statistics, and machine learning. 

►  Dr. AMY BRAVERMAN      

 —  Statistical Challenges for the Next Generation of NASA's Earth Observing Satellites Remote sensing data sets produced by NASA and other space agencies are a vast resource for the study of climate change and the physical processes that drive it. However, no remote sensing instrument actually observes these processes directly; instruments collect electromagnetic spectra aggregated over two-dimensional ground footprints or three-dimensional voxels (or sometimes just at a single point location). Inference on physical state based on these spectra occurs via complex, computationally intensive ground data processing algorithms. As we transition from the Earth Observing System (EOS, circa 1999-2025) to the new Earth System Observatory (ESO, circa 2026) data volumes will explode. For example, the Surface Biology and Geology (SBG) mission to be launched later this decade, will acquire millions of highly multivariate spectra per second over the global land surface of the Earth, at 30-60 meter resolution. This rich content of this vast information asset will remain mostly impenetrable unless new methods appropriate for these data and the questions they address, emerge. In this talk, I will give an overview the SBG mission, its scientific and societal goals, and three types of statistical challenges we face: science analysis, data processing, and uncertainty quantification. I will also describe our current thinking on how to address these challenges, and discuss opportunities for collaboration. 

Dr. Amy Braverman is a Senior Research Scientist at the Jet Propulsion Laboratory, California Institute of Technology. She holds a Ph.D.  in Statistics from UCLA, and came to JPL as a post-doctoral scholar in 1999. Prior to graduate school, she was a Research Director at Micronomics, Inc. in Los Angeles where she led teams preparing exhibits for complex civil litigation. Dr. Braverman worked on various NASA missions in various capacities over her 25 years at the Lab, first in designing data reduction methods for massive remote sensing data sets, and later expanding to address general statistical methodology and applications issues related to remote sensing. In 2012 she began working intensely on uncertainty quantification (UQ), and has developed practical methods for UQ in high-throughput, operational inverse problems of interest to NASA and JPL. She is now serving as the Chair of the SIAM Activity Group on Uncertainty Quantification, aiming to bridge the gap between traditional math-based UQ and statistics. Dr. Braverman is a Fellow of the American Statistical Association, and is the recipient of the NASA Exceptional Public Service Medal for her efforts to bring rigorous UQ to the NASA science enterprise. She especially enjoys working with post-docs, graduate students, and academic colleagues to solve new statistical research problems relevant to Earth and Space sciences. 

REGISTER 

→  Coffee & Refreshments at 3:30 pm 

→   Reception following the event.

STAMPS – Many problems in the physical sciences share common statistical challenges including heterogeneous data from multiple probes, uncertainty quantification, ill-posed inverse problems, spatio-temporal data and complex simulations. In 2018, a group of faculty and students at CMU started the STAMPS research group to develop new statistical and machine learning methodology tailored to the unique challenges that arise across multiple areas in the physical sciences. STAMPS provides foundational methodology in statistics, data science, machine learning and artificial intelligence for two distinct branches of physical science: (i) Astronomy and Particle Physics, and (ii) Climate and Environmental Science, which include applications in e.g. Oceanography, Meteorology, and Remote Sensing. STAMPS has become a vibrant forum for interdisciplinary exchange at the intersection of statistics and the physical sciences and will become a CMU Research Center in Fall 2024.

Event Website:
https://www.cmu.edu/dietrich/statistics-datascience/stamps/events/index.html#launch