We will be discussing ML techniques for generic density estimation with neural networks based on Masked Auto-Encoders (https://arxiv.org/abs/1502.03509, https://arxiv.org/abs/1705.07057), and an application to learning likelihood functions for inference (https://arxiv.org/abs/1805.07226). For context, these are generic and powerful tools, and some people are now proposing to use these methods for cosmological inference. Full details about this meeting will be posted here: https://www.benty-fields.com/manage_jc?groupid=191.
The Berkeley Statistics and Machine Learning Discussion Group meets weekly to discuss current applications across a wide variety of research domains and software methodologies. Register here to view, propose and vote for this group's upcoming discussion topics. All interested members of the UC Berkeley and LBL communities are welcome and encouraged to attend. Questions may be directed to François Lanusse.
Speaker(s)
François Lanusse
While at UC Berkeley, François Lanusse was a Data Science Fellow at BIDS and a Postdoctoral Scholar with the Berkeley Center for Cosmological Physics (BCCP) and the Foundations of Data Analysis (FODA) Institute, exploring the intersection between cosmology, statistics, and machine learning. His research was focused on measuring and exploiting the gravitational lensing effect (in which distant galaxies appear distorted due to the presence of massive structures along the line of sight) with the development of novel tools and methodologies based on sparse signal representations, convex optimization, and deep learning. Before joining Berkeley, Dr. Lanusse worked as a postdoctoral researcher within the McWilliams Center for Cosmology at Carnegie Mellon University, after completing a PhD in astrophysics at CEA Saclay near Paris.
Uroš Seljak
Uroš Seljak is a professor in UC Berkeley's Physics and Astronomy departments, a Senior Scientist at LBNL in the Physics Division, as well as a co-director of Berkeley Center for Cosmological Physics. His main research is in cosmology, where he combines theoretical, numerical, and data analysis methods to investigate the universe properties using cosmological observations, from cosmic microwave background to present day galaxy and dark matter distributions. His recent work combines statistics, numerical optimization, and N-body simulation methods to analyze large cosmological surveys, both space based (WMAP, Planck, Euclid, WFIRST) and ground based (SDSS, DESI, LSST). At Berkeley he teaches a course on Bayesian statistics and data science in Physics department. He has a PhD from MIT, was a postdoctoral fellow at Harvard and a faculty at Princeton University and Zurich University prior to Berkeley. He is a recipient of Sloan and Packard Fellowships, NSF CAREER award, and the Warner Prize of the American Astronomical Society.