2006 Spring Program on Astrostatistics

Research Foci

Historically, astronomy has served as fertile ground for stimulating the growth of new statistical and mathematical methodologies. Conversely, coping with the current and future needs of astronomy missions requires concerted efforts by cross-disciplinary collaborations involving astronomers, computer scientists, mathematicians and statisticians. This semester-long program at SAMSI provided a single geographical location - a crossroads - where researchers at the interface between statistics, applied mathematics, astronomy, and particle physics can congregate and initiate lasting collaborations.  Many of the research and educational activities will be joint with the Center for Astrostatistics at Penn State University.

Program Leaders: G. J. Babu (Penn State) - Chair, Eric Feigelson (Penn State), Tom Loredo (Cornell University), Donald Richards (Penn State), Alanna Connors (Eureka Scientific), and Larry Wasserman (Carnegie-Mellon University). Jim Berger (SAMSI) is Directorate Liaison, and Peter Bickel (Berkeley) is National Advisory Committee Liaison.

Description of Activities

Tutorials and Workshops

  • The Program began with Tutorials. Videos of most of the tutorials are also available, from 1/18/2006-1/22/2006, designed to familiarize statisticians with current trends in astronomy and expose astronomers to modern methodologies in statistics and applied mathematics.
  • The Opening Workshop was held 1/23/06-1/25/06, and focused on the scientific agenda of the program. It involved key leaders in astrostatistics, and was open to the community.
  • There were two Transition Workshops at the end of the program:

Working Groups

Working Groups met throughout the semester, focusing on research areas determined by the opening workshop.

  1. Exoplanets, with work on individual systems, populations studies, and optimal scheduling of observations; key statistics areas include inference with small samples & nonlinear models, MCMC algorithms for nonlinear/multimodel problems, hierarchical/empirical Bayesian methods, and experimental design.
  2. Surveys and Population Studies, including work on small or moderate scale surveys (exoplanets, Kuiper belt objects, GRB's), large-scale surveys (galaxy surveys, AGN surveys), modeling "Number-Size" or "Size-Frequency" distributions (power-law models, nonparametric models, comparison with large-scale numerical simulations), selection effects and source uncertainties (Malmquist-Eddington/Lutz-Kelker biases, handling upper limits, etc.), and coincidences between surveys; key statistics areas include survey sampling, regression and measurement error models/EVM, nonparametric modeling, survival analysis, and multiple testing.
  3. Gravitational Lensing, including work on the determination of basic statistics for random number of micro-images due to stars across a smooth dark matter background with external shear, computation of magnification probability distributions due to dark matter substructures, development of a statistical model of substructure population on galaxy-cluster scales, and exploration of the solution space for mass reconstruction.
  4. Source Detection and Feature Detection, including work on detecting point sources and extended sources in images, classification of detections, and anomaly detection.
  5. Particle Physics

Intensive Research Sessions

  1. Statistical Issues in Particle Physics: Louis Lyons (leader)
    March, with heavy emphasis during March 6-16
  2. Stellar Evolution: Bill Jefferys (leader)
    February 20-23
  3. Gravitational Wave Physics Workshop: Sam Finn (leader)
    Early April