J/ApJS/230/20 Machine learning technique to classify CoNFIG gal. (Aniyan+, 2017)
Classifying radio galaxies with the convolutional neural network.
Aniyan A.K., Thorat K.
<Astrophys. J. Suppl. Ser., 230, 20-20 (2017)>
=2017ApJS..230...20A 2017ApJS..230...20A (SIMBAD/NED BibCode)
ADC_Keywords: Galaxies, radio
Keywords: methods: miscellaneous; methods: observational;
radio continuum: galaxies; techniques: miscellaneous
Abstract:
We present the application of a deep machine learning technique to
classify radio images of extended sources on a morphological basis
using convolutional neural networks (CNN). In this study, we have
taken the case of the Fanaroff-Riley (FR) class of radio galaxies as
well as radio galaxies with bent-tailed morphology. We have used
archival data from the Very Large Array (VLA)-Faint Images of the
Radio Sky at Twenty Centimeters survey and existing visually
classified samples available in the literature to train a neural
network for morphological classification of these categories of radio
sources. Our training sample size for each of these categories is ∼200
sources, which has been augmented by rotated versions of the same. Our
study shows that CNNs can classify images of the FRI and FRII and
bent-tailed radio galaxies with high accuracy (maximum precision at
95%) using well-defined samples and a "fusion classifier," which
combines the results of binary classifications, while allowing for a
mechanism to find sources with unusual morphologies. The individual
precision is highest for bent-tailed radio galaxies at 95% and is 91%
and 75% for the FRI and FRII classes, respectively, whereas the recall
is highest for FRI and FRIIs at 91% each, while the bent-tailed class
has a recall of 79%. These results show that our results are
comparable to that of manual classification, while being much faster.
Finally, we discuss the computational and data-related challenges
associated with the morphological classification of radio galaxies
with CNNs.
Description:
We initially selected the FRI-II sample from a subset of the Combined
NVSS and FIRST Galaxies sample (CoNFIG henceforth; Gendre & Wall 2008,
J/MNRAS/390/819; Gendre+ 2010, J/MNRAS/404/1719).
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
table5.dat 68 187 Table of predictions for validation samples
--------------------------------------------------------------------------------
See also:
VIII/65 : 1.4GHz NRAO VLA Sky Survey (NVSS) (Condon+ 1998)
VIII/92 : The FIRST Survey Catalog, Version 2014Dec17 (Helfand+ 2015)
J/MNRAS/390/819 : Combined NVSS-FIRST Galaxies (CoNFIG) sample (Gendre+, 2008)
J/MNRAS/404/1719 : CoNFIG sample II (Gendre+, 2010)
J/ApJS/194/31 : Morphology for groups in the FIRST database (Proctor, 2011)
J/MNRAS/421/1569 : Properties of 18286 SDSS radio galaxies (Best+, 2012)
J/MNRAS/430/3086 : CoNFIG AGN sample (Gendre+, 2013)
J/MNRAS/446/2985 : Double-lobed radio sources catalog (van Velzen+, 2015)
Byte-by-byte Description of file: table5.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 19 A19 --- Name Source Name
21- 22 I2 h RAh Hour of Right Ascension (J2000)
24- 25 I2 min RAm Minute of Right Ascension (J2000)
27- 32 F6.3 s RAs Second of Right Ascension (J2000)
34 A1 --- DE- Sign of the Declination (J2000)
35- 36 I2 deg DEd Degree of Declination (J2000)
38- 39 I2 arcmin DEm Arcminute of Declination (J2000)
41- 45 F5.2 arcsec DEs Arcsecond of Declination (J2000)
47- 50 A4 --- True True Class of Source (1)
52- 58 A7 --- Pred Prediction by algorithm (1)
60- 68 F9.5 --- Prob Probability Score of Prediction (1)
--------------------------------------------------------------------------------
Note (1): This table contains the prediction results for the validation
samples from Config samples. The true class and the prediction by the
fusion classifier is given along with the probability scores.
BT = Bent-tailed (77 true; 64 predicted)
FRI = Fanaroff-Riley (FR) I (53 true; 53 predicted)
FRII = Fanaroff-Riley (FR) II (57 true; 69 predicted)
--------------------------------------------------------------------------------
History:
From electronic version of the journal
(End) Prepared by [AAS], Emmanuelle Perret [CDS] 07-Aug-2017