This is a summary of the `Astronomy Perspective of the 4th meeting on Statistical Challenges in Modern Astronomy held at Penn State University in June 2006. We comment on trends in the Astronomy community towards Bayesian methods and model selection criteria. We describe two examples where Bayesian methods have improved our inference: (i) photometric redshift estimation (ii) orbital parameters of extra-solar planets. We also comment on the pros and cons of Globalization of scientific research. Communities like Astronomy, High Energy Physics and Statistics develop ideas separately, but also have frequent interaction. This illustrates the benefits of comparing notes.
Download