Machine learning for transient recognition in difference imaging with minimum sampling effort

The amount of observational data produced by time-domain astronomy is exponentially in-creasing. Human inspection alone is not an effective way to identify genuine transients fromthe data. An automatic real-bogus classifier is needed and machine learning techniques are commonly used to achieve this goal. Building a training set with a sufficiently large number of verified transients is challenging, due to the requirement of human verification. We presentan approach for creating a training set by using all detections in the science images to be thesample of real detections and all detections in the difference images, which are generated by the process of difference imaging to detect transients, to be the samples of bogus detections. This strategy effectively minimizes the labour involved in the data labelling for supervised machine learning methods. We demonstrate the utility of the training set by using it to train several classifiers utilizing as the feature representation the normalized pixel values in 21-by-21pixel stamps centered at the detection position, observed with the Gravitational-wave Optical Transient Observer (GOTO) prototype. The real-bogus classifier trained with this strategy can provide up to 95% prediction accuracy on the real detections at a false alarm rate of 1%.

[1]  Tao Xiong,et al.  A combined SVM and LDA approach for classification , 2005, Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005..

[2]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[3]  G. A. Croes,et al.  FITS++: An Object-Oriented Set of C++ Classes to Support FITS , 1997 .

[4]  Larry Denneau,et al.  The Pan-STARRS wide-field optical/NIR imaging survey , 2010, Astronomical Telescopes + Instrumentation.

[5]  B. A. Boom,et al.  ScholarWorks @ UTRGV ScholarWorks @ UTRGV Properties of the Binary Black Hole Merger GW150914 Properties of the Binary Black Hole Merger GW150914 , 2016 .

[6]  Pablo A. Estévez,et al.  Supernovae detection by using convolutional neural networks , 2016, 2016 International Joint Conference on Neural Networks (IJCNN).

[7]  Eduardo Serrano,et al.  LSST: From Science Drivers to Reference Design and Anticipated Data Products , 2008, The Astrophysical Journal.

[8]  Umaa Rebbapragada,et al.  The Zwicky Transient Facility: Data Processing, Products, and Archive , 2018, Publications of the Astronomical Society of the Pacific.

[9]  A. Banday,et al.  Mining the Sky , 2001 .

[10]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[11]  Doug Tody,et al.  The Iraf Data Reduction And Analysis System , 1986, Astronomical Telescopes and Instrumentation.

[12]  Brad E. Tucker,et al.  Convolutional neural networks for transient candidate vetting in large-scale surveys , 2017, 1708.08947.

[13]  Evert Rol,et al.  A telescope control and scheduling system for the Gravitational-wave Optical Transient Observer (GOTO) , 2018, Astronomical Telescopes + Instrumentation.

[14]  Fang Yuan,et al.  SkyMapper Southern Survey: First Data Release (DR1) , 2018, Publications of the Astronomical Society of Australia.

[15]  The bulletin of mathematical biophysics , 2005, Protoplasma.

[16]  John S. Lewis Mining the Sky , 1996 .

[17]  Ralf Bender,et al.  Astronomical Data Analysis Software and Systems XVI ASP Conference Series , 2007 .

[18]  E. O. Ofek,et al.  Automating Discovery and Classification of Transients and Variable Stars in the Synoptic Survey Era , 2011, 1106.5491.

[19]  B. Metzger,et al.  Kilonovae , 2016, Living Reviews in Relativity.

[20]  D. Tody,et al.  IRAF in the Nineties , 1992 .

[21]  E. Bertin,et al.  SExtractor: Software for source extraction , 1996 .

[22]  B. Stalder,et al.  ATLAS: A High-cadence All-sky Survey System , 2018, 1802.00879.

[23]  A. J. Drake,et al.  FIRST RESULTS FROM THE CATALINA REAL-TIME TRANSIENT SURVEY , 2008, 0809.1394.

[24]  Pablo A. Estévez,et al.  Deep-HiTS: Rotation Invariant Convolutional Neural Network for Transient Detection , 2017, ArXiv.

[25]  B. A. Boom,et al.  GW170817: Observation of Gravitational Waves from a Binary Neutron Star Inspiral. , 2017, Physical review letters.

[26]  M. Wainwright,et al.  Using machine learning for discovery in synoptic survey imaging data , 2012, 1209.3775.

[27]  J. Kaplan,et al.  THE SLOAN DIGITAL SKY SURVEY-II SUPERNOVA SURVEY: TECHNICAL SUMMARY , 2007, 0708.2749.

[28]  S. Smartt,et al.  A First Catalog of Variable Stars Measured by the Asteroid Terrestrial-impact Last Alert System (ATLAS) , 2018, The Astronomical Journal.

[29]  J. Prochaska,et al.  Swope Supernova Survey 2017a (SSS17a), the optical counterpart to a gravitational wave source , 2017, Science.

[30]  R. Kotak,et al.  Machine learning for transient discovery in Pan-STARRS1 difference imaging , 2015, 1501.05470.

[31]  Emmanuel Bertin,et al.  Mining Pixels: The Extraction and Classification of Astronomical Sources , 2001 .