Nonparametric Inference for Complex Data and Models in Astronomy

• Activity Number: 335
• Type: Invited
• Date/Time: Tuesday, August 11, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Physical and Engineering Sciences
• Author(s): Ann Lee*
• Companies: Carnegie Mellon University
• Keywords: high-dimensional inference ; spectral methods ; astrostatistics ; approximate likelihoods ; kernel machine learning ; nonparametric

Abstract

Many estimation problems in astronomy are highly complex with non-standard aggregate data objects (e.g., images, spectra, and light curves) which are not directly amenable to traditional nonparametric statistical analysis. At the same time, the physical processes that generated these data may themselves be complicated; often the only existing "theory" is in the form of complex simulations. In this talk, I will describe our efforts toward developing nonparametric methods that fully exploit massive data rich in information but complex in structure, and our work toward modeling such data beyond regression and classification. (Part of this work is joint with Peter Freeman, Rafael Izbicki, and Chad Schafer.)