The 3rd edition of INPE's Advanced School in Astrophysics has Astrostatistics as its theme. Simply put, Astrostatistics deals with particularities of Statistics in the realm of Astronomy: huge amounts of data, a variety of data coming from different parts of the electromagnetic spectrum (and even outside it), different forms of data and applications demanding responses to real-time events.

As in so many other fields of research, Astrostatistics progressed together with the advances in the numerical computing technologies. The free-fall trajectory of prices of computers in the last 20 years, particularly the personal ones, made possible calculations in seconds that once took thousands of brain-hours of work in camp-sized facilities. The spread in the use of computer power is still in its beginnings if we observe the world as a whole. In the astronomical context, the recent experience of virtual observatories adds new challenges and opportunities to the context of Astrostatistics.

The 3rd INPE Advanced School in Astrophysics offers the Brazilian astronomical community (specially PhD students) an excellent opportunity of exposition to the latest tools developed both in the classical context of statistics as in the one that makes extensive use of simulations to model physical phenomena.

The above statistics approaches applied to astrophysics will be addressed at INPE's III Advanced Course on Astrophysics by a team of renowned lecturers:

Dr. Thomas Loredo combine data analysis with theoretical astrophysics testing astronomical models and theories using Bayesian statistics, particularly in high energy astrophysics and cosmology. His work focuses on problems that can benefit from development of new statistical methodology, and largely adopts the Bayesian approach to statistics. Recently, his work has also addressed statistical issues arising in the study of extrasolar planets and analysis of the distribution of trans-Neptunian objects (including Kuiper belt objects). Tom has been the principal investigator for a NASA-sponsored project developing a statistical inference package using the Python computing language. Dr. Loredo has been a lecturer in the “Summer School in Statistics for Astronomers & Physicists” organized by the “Center for Astrostatistics” of the Pennsylvania State University.

Hedibert Freitas Lopes conducts research in Markov Chain Monte Carlo techniques and Sequential Monte Carlo methods applied, for example, to time-series models; modeling time-varying covariance of multivariate time series through latent factor analysis; Choleski decomposition and other factorizations; dynamic models and Bayesian inference; and computation. He is mainly interested in the implementation of the Bayesian paradigm to solve real large-scale problems. His research highlights the importance of model uncertainty and how it can be measured and accounted for based on modern computational statistics schemes.

Dr. Eric Feigelson wrote his dissertation at Harvard University under the supervision of Nobel Prize winner Riccardo Giacconi. He joined the Astronomy & Astrophysics faculty at Penn State in 1983 where he is now Professor. He has studied various topics in X-ray astronomy, focusing on its implications for the formation and early evolution of stars and planets. He also has a long-standing collaboration with statisticians to improve astronomical data analysis. He was a John Parker Fellow at Harvard, shared two NASA Group Achievement Awards and Rossi Prize, received an NSF Presidential Young Investigator Award, and a Department certificate for distinguished teaching. He has led the astronomy undergraduate and public outreach programs at Penn State, and co-founded its Center for Astrostatistics. He has coauthored over 170 refereed articles and 5 books in X-ray astronomy and statistics. Dr. Feigelson has been also a lecturer in the "Summer School in Statistics for Astronomers & Physicists" organized by the Center for Astrostatistics at the Pennsylvania State University.