The Statistical and Applied Mathematical Sciences Institute, located in the Research Triangle of North Carolina, will host a year-long program entitled "Statistical, Mathematical and Computational Methods for Astronomy." Among a handful of planned working groups is one focused on LSST statistical challenges. Specific research areas are under discussion; please contact the organizers if you have suggestions.
Informatics & Statistics Science Collaboration Activities
Statistical Challenges in Modern Astronomy VI will be held at Carnegie Mellon University, June 6 through 10, 2016. This meeting will continue the interdisciplinary tradition of its predecessors, bringing together researchers in astronomy, cosmology, statistics, and machine learning to facilitate progress on the significant data analysis challenges that result from current and future astronomical sky surveys. The meeting is co-chaired by Shirley Ho and Chad Schafer, and will incorporate significant LSST-related content.
The importance of likelihood-free methods of inference becomes evident when considering the complexity of deriving a sufficiently accurate approximation to the likelihood function in astronomical problems. For example: How does one adequately incorporate the effect of photometric redshifts when assumptions of Gaussian-distributed errors are unrealistic? ISSC members are working with other science collaborations to develop and test methods of approximate Bayesian computation (ABC) for use in LSST inference challenges.
Synoptic time-domain surveys provide astronomers, not simply more data, but a different kind of data: large ensembles of multivariate, irregularly and asynchronously sampled light curves. We are building a statistical framework for light curve demography—optimal accumulation and extraction of information, not only along individual light curves as conventional methods do, but also across large ensembles of related light curves.
Hosted by the Center for Astrostatistics at Penn State University since 2005, the annual Summer School in Statistics for Astronomers is designed for graduate students and researchers in astronomy. The school is an intensive week covering basic statistical inference, several fields of applied statistics, and the R computing environment. A repertoire of well-established techniques applicable to observational astronomy and physics are developed.