Title: Nonparametric Tools for Complex Data and Simulations
Ann B. Lee, Department of Statistics & Data Science, Carnegie Mellon University
Abstract:
Recent technological advances have led to a rapid growth in not just the amount of scientific data but also their complexity and richness. Simulation models have, at the same time, become increasingly detailed and better at capturing the underlying processes that generate observable data. On the statistical methods front, however, we still lack tools that accurately quantify complex relationships between data and model parameters, as well as adequate tools to validate approximate likelihoods and emulators fit to high-dimensional data and computationally intensive simulations. In this talk, I will discuss our current work on how we leverage machine learning methods toward some of these statistical inference challenges. I will draw examples from astronomy on photometric redshift estimation and cosmological parameter inference. (Part of this work is joint with Rafael Izbicki, Taylor Pospisil, Peter Freeman, Niccolo Dalmasso, Ilmun Kim, and the LSST-DESC-PZ working group)
Bio:
Ann Lee is an associate professor in the department of statistics and data science and the machine learning department at Carnegie Mellon University. She received her doctorate in physics at Brown University, and was prior to joining CMU in 2015, the J.W. Gibbs Assistant Professor at Yale University’s Program of Applied Mathematics. Ann Lee’s interests are in developing statistical methodology for the type of complex data and problems often encountered in the physical sciences. She is particularly interested in scalable methods that adapt to the intrinsic structure of data, and nonparametric approaches that can handle different types of data, such as data arriving from multiple scientific probes. Her recent work includes conditional density estimation in a high-dimensional regression setting, faster alternatives to approximate Bayesian computing, and applications in astronomy and hurricane intensity guidance involving satellite imagery and massive astronomical surveys.