Artificial neural networks for selection of pulsar candidates from radio continuum surveys

Pulsar search with time-domain observation is very computationally expensive and data volume will be enormous with the next generation telescopes such as the Square Kilometre Array. We apply artificial neural networks (ANNs), a machine learning method, for efficient selection of pulsar candidates from radio continuum surveys, which are much cheaper than time-domain observation. With observed quantities such as radio fluxes, sky position and compactness as inputs, our ANNs output the "score" that indicates the degree of likeliness of an object to be a pulsar. We demonstrate ANNs based on existing survey data by the TIFR GMRT Sky Survey (TGSS) and the NRAO VLA Sky Survey (NVSS) and test their performance. Precision, which is the ratio of the number of pulsars classified correctly as pulsars to that of any objects classified as pulsars, is about 96$\%$. Finally, we apply the trained ANNs to unidentified radio sources and our fiducial ANN with five inputs (the galactic longitude and latitude, the TGSS and NVSS fluxes and compactness) generates 2,436 pulsar candidates from 456,866 unidentified radio sources. These candidates need to be confirmed if they are truly pulsars by time-domain observations. More information such as polarization will narrow the candidates down further.

[1]  A. Fontana,et al.  Photometric redshifts with the Multilayer Perceptron Neural Network: Application to the HDF-S and SDSS , 2003, astro-ph/0312064.

[2]  R. Manchester,et al.  The ATNF Pulsar Catalogue , 2003, astro-ph/0309219.

[3]  D. Frail,et al.  A radio spectral index map and catalogue at 147-1400 MHz covering 80 per cent of the sky , 2017, 1711.11367.

[4]  Ofer Lahav,et al.  ANNz: Estimating Photometric Redshifts Using Artificial Neural Networks , 2004 .

[5]  R. N. Manchester,et al.  Tests of General Relativity from Timing the Double Pulsar , 2006, Science.

[6]  F. Camilo,et al.  The Parkes multi-beam pulsar survey - I. Observing and data analysis systems, discovery and timing of 100 pulsars , 2001, astro-ph/0106522.

[7]  R. P. Eatough,et al.  Selection of radio pulsar candidates using artificial neural networks , 2010, 1005.5068.

[8]  Joeri van Leeuwen,et al.  Applying Deep Learning to Fast Radio Burst Classification , 2018, The Astronomical Journal.

[9]  D. Stinebring,et al.  Gravitational Wave Astronomy Using Pulsars: Massive Black Hole Mergers & the Early Universe , 2009, 0902.2968.

[10]  A. J. Faulkner,et al.  Pulsar searches and timing with the square kilometre array , 2009 .

[11]  Hui Zhang,et al.  Pulsar candidate selection using ensemble networks for FAST drift-scan survey , 2019, Science China Physics, Mechanics & Astronomy.

[12]  Marco O. P. Sampaio,et al.  Testing general relativity with present and future astrophysical observations , 2015, 1501.07274.

[13]  B. Reid,et al.  Arecibo Pulsar Survey Using ALFA. I. Survey Strategy and First Discoveries , 2005, astro-ph/0509732.

[14]  Steve Croft,et al.  Low-Frequency Spectral Energy Distributions of Radio Pulsars Detected with the Murchison Widefield Array , 2017, Publications of the Astronomical Society of Australia.

[15]  C. Lintott,et al.  Galaxy Zoo: reproducing galaxy morphologies via machine learning★ , 2009, 0908.2033.

[16]  et al,et al.  Optical and Radio Properties of Extragalactic Sources Observed by the FIRST Survey and the Sloan Digital Sky Survey , 2002, astro-ph/0202408.

[17]  A. Hewish,et al.  Observation of a Rapidly Pulsating Radio Source , 1968, Nature.

[18]  E. Flesch.  The Half Million Quasars (HMQ) Catalogue , 2015, Publications of the Astronomical Society of Australia.

[19]  The Parkes Pulsar Timing Array Project , 2012, Publications of the Astronomical Society of Australia.

[20]  O. Lahav,et al.  An artificial neural network approach to the classification of galaxy spectra , 1996, astro-ph/9608073.

[21]  B. C. Joshi,et al.  A Search for Pulsars in Steep Spectrum Radio Sources , 2018, The Astrophysical Journal.

[22]  Kurt Hornik,et al.  Approximation capabilities of multilayer feedforward networks , 1991, Neural Networks.

[23]  Shanoli Samui Pal,et al.  Photo-$z$ with CuBAN$z$: An improved photometric redshift estimator using Clustering aided Back Propagation Neural network , 2016, 1609.03568.

[24]  E. Greisen,et al.  The NRAO VLA Sky Survey , 1996 .

[25]  R. P. Eatough,et al.  A Cosmic Census of Radio Pulsars with the SKA , 2014, 1501.00056.

[26]  B. Semelin,et al.  Analysing the 21 cm signal from the epoch of reionization with artificial neural networks , 2017, 1701.07026.

[27]  A. Naim,et al.  Automated morphological classification of APM galaxies by supervised artificial neural networks , 1995, astro-ph/9503001.

[28]  D. Frail,et al.  An image-based search for pulsars among Fermi unassociated LAT sources , 2017, 1712.06609.

[29]  O. Lahav,et al.  Morphological Classification of galaxies by Artificial Neural Networks , 1992 .

[30]  D. Backer,et al.  Constructing a Pulsar Timing Array , 1990 .

[31]  D. Schnitzeler,et al.  Modelling the Galactic distribution of free electrons , 2012, 1208.3045.

[32]  C. Flynn,et al.  SPINN: a straightforward machine learning solution to the pulsar candidate selection problem , 2014, 1406.3627.

[33]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[34]  J. Han,et al.  The Spatial Energy Spectrum of Magnetic Fields in Our Galaxy , 2004, astro-ph/0404221.

[35]  R. Manchester,et al.  The Australia Telescope National Facility Pulsar Catalogue , 2005 .

[36]  R. Ekers,et al.  Detecting pulsars with interstellar scintillation in variance images , 2016, 1607.07740.

[37]  D. Frail,et al.  The GMRT 150 MHz all-sky radio survey - First alternative data release TGSS ADR1 , 2016, 1603.04368.

[38]  D. A. García-Hernández,et al.  University of Birmingham The Fourteenth Data Release of the Sloan Digital Sky Survey: , 2017 .

[39]  George Cybenko,et al.  Approximation by superpositions of a sigmoidal function , 1992, Math. Control. Signals Syst..

[40]  Ping Guo,et al.  Pulsar Candidate Identification with Artificial Intelligence Techniques , 2017, ArXiv.

[41]  S. Desai,et al.  Separation of pulsar signals from noise using supervised machine learning algorithms , 2017, Astron. Comput..

[42]  J. Lattimer,et al.  The Physics of Neutron Stars , 2004, Science.

[43]  D. Champion,et al.  The European Pulsar Timing Array and the Large European Array for Pulsars , 2013 .

[44]  F. Jankowski,et al.  Spectral properties of 441 radio pulsars , 2017, 1709.08864.

[45]  G. Desvignes,et al.  SEARCHING FOR PULSARS USING IMAGE PATTERN RECOGNITION , 2013, 1309.0776.

[46]  H. Falcke,et al.  A LOFAR census of non-recycled pulsars: average profiles, dispersion measures, flux densities, and spectra , 2015, 1511.01767.

[47]  Daniel George,et al.  Deep Learning for Real-time Gravitational Wave Detection and Parameter Estimation with Advanced LIGO Data , 2017, ArXiv.