Table Of Content

Springer Series in (cid:65)(cid:115)(cid:116)(cid:114)(cid:111)(cid:115)(cid:116)(cid:97)(cid:116)(cid:105)(cid:115)(cid:116)(cid:105)(cid:99)(cid:115) Editor-in-chief: Joseph M. Hilbe, Jet Propulsion Laboratory, and Arizona State University, USA Jogesh Babu, The Pennsylvania State University, USA Bruce Bassett, University of Cape Town, South Africa Steffen Lautitzen, Oxford University, UK Thomas Loredo, Cornell University, USA Oleg Malkov, Moscow State University, Russia Jean-luc Starck, CEA/Saclay, France David van Dyk, Imperial College, London, UK Springer Series in Astrostatistics, #1 Forfurther volumes: http://www.springer.com/series/1432 Joseph M. Hilbe Editor Astrostatistical Challenges for the New Astronomy Editor Joseph M. Hilbe Arizona State University Tempe, AZ, USA ISBN 978-1-4614-3507-5 ISBN 978-1-4614-3508-2 (eBook) DOI 10.1007/978-1-4614-3508-2 Springer New York Heidelberg Dordrecht London Library of Congress Control Number: 2012948130 © Springer Science+Business Media New York 2013 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com) Preface Although statistical analysis was first applied to agricultural, business, and architectural data as well as to astronomical problems at least thirty to forty centuries ago, astronomers have largely been estranged from statistics throughout the past two centuries. Except for a few astronomers, the majority have paid little attention to the advances made in statistical theory and application since the work of Laplace and Gauss in the late 18th and early 19th centuries. Descriptive statistical analysis, of course, has continued to be used throughout this period, but inferential methods were largely ignored. It was not until the last decades of the 20th century that computing speed and memory allowed the development of statistical software that was rigorous enough to entice astronomers to again become interested in inferential statistics. During this time small groups of astronomers and statisticians joined together developing both collaborations and conferences to discuss statistical methodology as it applies to astronomical data. There are a variety of reasons for this regained interest, as well as for the two- century breach. The initial monograph in the text addresses the reasons for the separation, and for the renewed interest. Newly developed instruments are - and will be - gathering truly huge amounts astronomical data which will need to be statistically evaluated and modeled. Much of the data being generated appears to be complex, and not amenable to straightforward traditional methods of analysis. Many astrostatisticians -- astronomers, statisticians and computation theorists having an interest in the statistical analysis of astronomical/cosmological data -- have recognized the inherent multifarious nature of astronomical events, and have recently turned to Bayesian methods for analyzing astronomical data, abandoning at least in part the traditional frequentist methods that characterized the discipline during the 19th and 20th centuries. However, frequentist-based analyses still have an important role in astrostatistics, as will be learned from this volume. Except for the first two introductory chapters, the collection of monographs in this volume present an overview of how leading astrostatisticians are currently dealing with astronomical data. The initial article is by the editor, providing a brief history of astrostatistics, and an overview of its current state of development. It is an adaptation of a presentation delivered at Los Alamos National Laboratory in October, 2011 as a seminar in the Information Science and Technology Center v vi Preface seminar series. The second contribution is by Thomas Loredo of Cornell University who presents an overview of both frequentist and Bayesian methodologies and details how each has been misunderstood. He then outlines considerations for future astrostatistical analyses. Prof Loredo's contribution is an adaptation of a presentation he made on this topic at the Statistical Challenges in Modern Astronomy V conference held at Pennsylvania State University in June, 2011. The other monographs describe ongoing state-of-the-art astrostatistical research. Contributions to this volume are largely adaptations of selected invited and special topics session presentations given by the authors at the 2011 ISI World Statistics Congress in Dublin, Ireland. Several monographs have been completely re-written from their original form for this volume, and some are entirely new. Nearly all of the contributors as well as co-authors are recognized as leading -- if not the leading -- astrostatisticians in the area of their research and contribution to this book. The value of a volume such as this rests in the fact that readers get the opportunity to review the work of top astrostatisticians. Each selection employs a different manner of applying statistical methods to the data being evaluated. Articles using traditional methods utilize techniques that are common among statisticians who model astronomical data. Together with a history and overview of astrostatistics as a discipline, the research shared in this volume will provide readers with a good sense of the current state of astrostatistics. The articles may also encourage readers themselves to contribute to this new area of statistics. The challenges are great, and may involve developing new methods of statistical analysis. The answers that may be gained, though, address questions that are central to astronomy and cosmology. In fact, substantial statistical work has already been done in such astrophysical areas as high-energy astronomy (e.g., X- ray, gamma, and cosmic ray astronomy), neutrino astrophysics, image analysis, and both extra-solar and early galaxy formation. Even elusive areas such as understanding dark matter and dark energy, and if a multiverse exists, are queries that may ultimately be answered using novel applications of current statistical functions, or they may require the application of new and innovative astrostatistical methods yet to be developed. It will be clear on reviewing the monographs in this text that astrostatistics is a discipline coming to its own. It is involved with evaluating problems of cosmic concern, and as it advances it will demand the most of both computational and statistical resources. I wish to thank Marc Strauss, Springer statistics editor, for suggesting the initiation of a Springer Series in Astrostatistics, and for his support for this text. Without his assistance this book would never have been constructed. I also must acknowledge Rajiv Monsurate for his fine work in setting up the book for publication, and Hannah Bracken, for her editorial support. Their efforts were also essential for this book's completion. Finally I dedicate this book to my late father, Rader J Hilbe, who fostered my interest in both mathematics and astronomy. Preface vii Update: On August 30, 2012 the Executive Board of the ISI Astrostatistics Network approved its being re-organized as the International Astrostatistics Association (IAA), the first professional association for the discipline of astrostatistics. The initial officers of the IAA were also approved. Membership includes researchers from astronomy/astrophysics, physics, statistics, the computer-information sciences and others who are interested in the statistical analysis of astronomical data. The IAA has three classes of membership, with no difference in benefits: regular, Post Doc, and student. As the professional association for the discipline, the IAA fosters the creation of astrostatistics committees and working groups within both international and national astronomical and statistical organizations. During the same week that the IAA was created, the International Astronomical Union approved an IAU Astrostatistics Working Group. The IAA, IAU Astrostatistics WG, ISI astrostatistics committee, and LSST share a common Astrostatistics and AstroInformatics Portal web site: https://asaip.psu.edu The IAA and ISI are co-sponsors of the Springer Series in Astrostatistics. Contents Preface ...................................................................................................................... v 1 Astrostatistics: A brief history and view to the future .................................. 1 Joseph M. Hilbe, Jet Propulsion Laboratory and Arizona State University 2 Bayesian astrostatistics: A backward look to the future. ............................ 15 Thomas J. Loredo, Dept of Astronomy, Cornell University 3 Understanding better (some) astronomical data using Bayesian methods ............................................................................................................ 41 S. Andreon, INAF-Osservatorio Astronomico di Brera 4 BEAMS: separating the wheat from the chaff in supernova analysis ............................................................................................................. 63 Martin Kunz, Institute for Theoretical Physics, Univ of Geneva, Switz; Renée Hlozek, Dept of Astrophysical Sciences, Princeton Univ, NJ; Bruce A. Bassett, Dept of Mathematics and Applied Mathematics, Univ of Cape Town, SA; Mathew Smith, Dept of Physics, Univ of Western Cape, SA; James Newling, Dept of Mathematics and Applied Mathematics, Univ of Cape Town, SA; Melvin Varughese, Dept of Statistical Sciences, Univ of Cape Town, SA 5 Gaussian Random Fields in Cosmostatistics. ............................................... 87 Benjamin D. Wandelt, Institut d'Astrophysique de Paris, Université Pierre et Marie Curie, France 6 Recent advances in Bayesian inference in cosmology and astroparticle physics thanks to the Multinest Algorithm .................. 107 Roberto Trotta, Astrophysics Group, Dept of Physics, Imperial College London; Farhan Feroz (Cambridge), Mike Hobson (Cambridge), and Roberto Ruiz de Austri (Univ of Valencia, Spain) 7 Extra-solar Planets via Bayesian Fusion MCMC. ..................................... 121 Philip C. Gregory, Dept of Astronomy, University of British Columbia, Canada ix x Contents 8 Classification and Anomaly Detection for Astronomical Survey Data ................................................................................................... 149 Marc Henrion, Dept of Mathematics, Imperial College, London, UK; Daniel J. Mortlock (Imperial), David J. Hand (Imperial), and ) Axel Gandy (Imperial)) 9 Independent Component Analysis for dimension reduction classification: Hough transform and CASH Algorithm. ........................... 185 Asis Kumar Chattopadhyay, Dept of Statistics, University of Calcutta, India; Tanuka Chattyopadhyay (Univ Calcutta), Tuli De (Univ Calcutta), and Saptarshi Mondal (Univ Calcutta) 10 Improved cosmological constraints from a Bayesian hierarchical model of supernova type Ia data ................................................................. 203 Marisa Cristina March, (University of Sussex); Roberto Trotta (Imperial), Pietro Berkes (Brandeis Univ), Glenn Starkman (Case Western Reserve Univ), Pascal Vaudrevange (DESY, Hamburg, Germany)