J/ApJS/268/28 LAMOST DR9 white dwarf cand. from deep-learning (Tan+, 2023)
A white dwarf search model based on a deep transfer-learning method.
Tan L., Liu Z., Wang F., Mei Y., Deng H., Liu C.
<Astrophys. J. Suppl. Ser., 268, 28 (2023)>
=2023ApJS..268...28T 2023ApJS..268...28T
ADC_Keywords: Stars, white dwarf; Spectra, optical; Models
Keywords: Convolutional neural networks ; White dwarf stars ;
Stellar classification ; Stellar types
Abstract:
White dwarfs represent the ultimate stage of evolution for over 97% of
stars and play a crucial role in studies of the Milky Way's structure
and evolution. Recent years have witnessed significant progress in
using deep-learning methods for identifying unique objects in
large-scale data. In this paper, we present a model based on transfer
learning for identifying white dwarfs. We constructed a data set using
the spectra released by LAMOST DR9 and trained a convolutional neural
network model. The model was then further trained using a
transfer-learning approach for a binary classification model. Our
final model is comprised of a seven-class classification model and a
binary classification model. The testing set yielded an accuracy rate
of 96.08%. Our proposed model successfully identifies 4314 of the
4479 white dwarfs published in previous papers. We applied this model
to filter the 1,121,128 spectral data from the LAMOST DR9 V1 catalog.
Subsequently, we obtained 6317 white dwarf candidates, of which
5014 were cross-validated and found to be known white dwarfs. We
finally identified 489 new white dwarfs out of the remaining
1303 candidates, containing 377 DAs, 1 DB, 4 DZs, 1 magnetic WD,
101 DA+M binaries, and 1 DB+M binary. Our study also compared
transfer-learning methods with non-transfer-learning methods, and the
results show that transfer learning provides faster training speed and
a higher accuracy rate.
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
table4.dat 150 6317 Manual verification results of the
6317 white dwarf candidates
table5.dat 149 4432 Basic information of 4432 white dwarf candidates
--------------------------------------------------------------------------------
See also:
III/129 : Spectroscopically Identified White Dwarfs (McCook+, 1987)
III/210 : Spectroscopically Identified White Dwarfs (McCook+, 1999)
III/235 : Spectroscopically Identified White Dwarfs (McCook+, 2008)
V/156 : LAMOST DR7 catalogs (Luo+, 2019)
J/ApJ/607/426 : Catalog of white dwarfs in SDSS-DR1 (Kleinman+, 2004)
J/ApJS/204/5 : SDSS DR7 white dwarf catalog (Kleinman+, 2013)
J/AJ/146/34 : IR photometry of DA white dwarfs from LAMOST (Zhang+, 2013)
J/MNRAS/452/765 : WD candidates using LAMOST DR3 (Gentile Fusillo+, 2015)
J/MNRAS/454/2787 : LAMOST DR2 white dwarfs (Guo+, 2015)
J/MNRAS/446/4078 : New white dwarf stars in SDSS DR10 (Kepler+, 2015)
J/MNRAS/455/3413 : White dwarf and subdwarf stars in SDSS DR12 (Kepler+, 2016)
J/MNRAS/479/415 : Galaxy mergers sample (Ackermann+, 2018)
J/MNRAS/480/4505 : A white dwarf cat. from Gaia-DR2 (Jimenez-Esteban+, 2018)
J/MNRAS/477/4641 : LAMOST DR5 White dwarf-main sequence binaries (Ren+, 2018)
J/MNRAS/482/4570 : Gaia DR2 white dwarf candidates (Gentile Fusillo+, 2019)
J/MNRAS/486/2169 : White dwarf and subdwarf stars in SDSS DR14 (Kepler+, 2019)
J/MNRAS/482/965 : Age of the Gal. stellar halo from Gaia WDs (Kilic+, 2019)
J/other/RAA/19.88 : DB white dwarfs from the LAMOST DR5 (Kong+, 2019)
J/MNRAS/508/3877 : Catalogue of white dwarfs in Gaia EDR3 (Gentile+, 2021)
J/ApJS/259/5 : Hot subdwarf candidates from LAMOST DR7 (Tan+, 2022)
Byte-by-byte Description of file: table4.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 19 A19 --- LAMOST LAMOST DR9 designation (JHHMMSS.ss+DDMMSS.s)
21- 30 F10.6 deg RAdeg LAMOST Right Ascension (J2000)
32- 40 F9.6 deg DEdeg [-8.3/86] LAMOST Declination (J2000)
42- 50 I9 --- ObsID [112211/914109128] LAMOST observation ID
52- 57 F6.2 --- gSNR [10/488] g-band spectral signal-to-noise ratio
59- 66 F8.2 --- muSNR [0/180]?=-9999 µ-band spectral
signal-to-noise ratio
68- 75 F8.2 --- zSNR [0/388]?=-9999 z-band spectral
signal-to-noise ratio
77- 84 F8.2 --- rSNR [0/437]?=-9999 r-band spectral
signal-to-noise ratio
86- 93 F8.2 --- iSNR [0/483]?=-9999 i-band spectral
signal-to-noise ratio
95- 111 A17 --- Source Results of cross-validation and
identification (G1)
113- 150 A38 --- Subclass White dwarf subclass (G2)
--------------------------------------------------------------------------------
Byte-by-byte Description of file: table5.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 19 A19 --- LAMOST LAMOST designation
21- 30 F10.6 deg RAdeg LAMOST Right Ascension (J2000)
32- 40 F9.6 deg DEdeg [-7.7/88] LAMOST Declination (J2000)
42- 50 I9 --- ObsID [108169/911408009] LAMOST observation ID
52- 55 F4.2 --- gSNR [5/9.99] g-band spectral signal-to-noise ratio
57- 64 F8.2 --- muSNR [0/16]?=-9999 µ-band spectral
signal-to-noise ratio
66- 73 F8.2 --- zSNR [0/138]?=-9999 z-band spectral
signal-to-noise ratio
75- 82 F8.2 --- rSNR [0/71]?=-9999 r-band spectral signal-to-noise
ratio
84- 91 F8.2 --- iSNR [0/116]?=-9999 i-band spectral
signal-to-noise ratio
93- 111 A19 --- Source Results of cross-validation and
identification (G1)
113- 149 A37 --- Subclass White dwarf subclass (G2)
--------------------------------------------------------------------------------
Global notes:
Note (G1): Results of cross-validation and identification as follows:
a = Zhao et al. 2013AJ....145..169Z 2013AJ....145..169Z
b = Ren et al. 2018, J/MNRAS/477/4641
c = Kong et al. 2019, J/other/RAA/19.88
d = Guo et al. 2015, J/MNRAS/454/2787
e = Guo et al. 2022MNRAS.509.2674G 2022MNRAS.509.2674G
f = Kong & Luo 2021RNAAS...5..249K 2021RNAAS...5..249K
g = Kepler et al. 2015, J/MNRAS/446/4078
g = Kepler et al. 2019, J/MNRAS/486/2169
i = Kepler et al. 2021MNRAS.507.4646K 2021MNRAS.507.4646K
j = Kleinman et al. 2004, J/ApJ/607/426
k = Kepler et al. 2007MNRAS.375.1315K 2007MNRAS.375.1315K
l = Kleinman et al. 2013, J/ApJS/204/5
m = Kepler et al. 2016, J/MNRAS/455/3413
n = Gentile Fusillo et al. 2015, J/MNRAS/452/765
o = Gentile Fusillo et al. 2021, J/MNRAS/508/3877
LAMOST = white dwarfs marked by the LAMOST pipeline
NEW = newly discovered white dwarfs (489 occurrences)
MIS = stars manually identified as other classes (814 occurrences)
CANDIDATE = white dwarf candidates (2287 occurrences)
Note (G2): As some of the samples have been validated multiple times, the
subclasses correspond to the sources (Column "Source") in turn.
--------------------------------------------------------------------------------
History:
From electronic version of the journal
(End) Prepared by [AAS], Emmanuelle Perret [CDS] 20-Dec-2023