J/AJ/166/189 Classif. of TESS LCs for T Tauri stars (Rodriguez-Feliciano+, 2023)
Machine-learning morphological classification of TESS light curves of T Tauri
stars.
Rodriguez-Feliciano B., Serna J., Garcia-Varela A., Hernandez J.,
Cabrera-Garcia J.F.
<Astron. J., 166, 189 (2023)>
=2023AJ....166..189R 2023AJ....166..189R
ADC_Keywords: YSOs; Stars, variable; Photometry; Optical
Keywords: Young stellar objects ; Light curve classification ;
Variable stars ; Astrostatistics techniques ; Protoplanetary disks
Abstract:
We present a variability and morphological classification study of
TESS light curves for TTauri star candidates in the Orion, IC348,
γVelorum, Upper Scorpius, Corona Australis, and Perseus OB2
regions. We propose 11 morphological classes linking brightness
variation behaviors with possible physical or geometric phenomena
present in TTauri stars, and develop a supervised machine-learning
algorithm to automate the classification among these. Our algorithm
optimizes and compares the true positive rate (recall) among k-nearest
neighbors, classification trees, random forests, and support vector
machines. This is done characterizing light curves with features
depending on time, periodicity, and magnitude distribution. Binary and
multiclass classifiers are trained and interpreted in a way that
allows our final algorithm to have single or mixed classes. In the
testing sample, the algorithm assigns mixed classes to 27% of the
stars, reaching up to five simultaneous classes. A catalog of 3672
TTauri star candidates is presented, along with their possible period
estimations, predicted morphological classes, and visually revised
ones. The cross-validation estimated performance of the final
classifiers is reported. Binary classifiers surpass multiclass recall
values for classes with less representation in the training sample.
Support vector machines and random forest classifiers obtain better
recalls. For comparison, another performance estimation of the final
classifiers is calculated using the revised classes of our testing
sample, indicating that this performance excels in singled classed
stars, which happens in about 75% of the testing sample.
Description:
We use the TESSExtractor tool (Serna+ 2021, J/ApJ/923/177) to perform
LCs from the TESS cycle 1 and 2 (30 minutes of cadence). This tool
downloads 10x10 pixel cutouts from the FFIs using the TESScut service,
selects an optimal aperture, and performs simple aperture photometry.
The flux was extracted using circular apertures from 1.0-5.0 pixels of
radius and properly calibrated to the TESS photometric system (See
TICv8.0, Cat. IV/38). See Section 2.
File Summary:
--------------------------------------------------------------------------------
FileName Lrecl Records Explanations
--------------------------------------------------------------------------------
ReadMe 80 . This file
table7.dat 92 3672 Catalog of both the training and testing samples
--------------------------------------------------------------------------------
See also:
II/328 : AllWISE Data Release (Cutri+ 2013)
IV/38 : TESS Input Catalog - v8.0 (TIC-8) (Stassun+, 2019)
IV/39 : TESS Input Catalog version 8.2 (TIC v8.2) (Paegert+, 2021)
J/AJ/129/907 : New Ori OB1 members (Briceno+, 2005)
J/A+A/463/1017 : B and V light curves of AA Tau (Bouvier+, 2007)
J/ApJ/733/50 : MIR photometric monitoring of ONC (Morales-Calderon+, 2011)
J/ApJ/758/31 : IR photometry for members of Upper Sco (Luhman+, 2012)
J/AJ/147/82 : Monitoring of disk-bearing stars in NGC 2264 (Cody+, 2014)
J/ApJ/794/36 : σOrionis cluster stellar population (Hernandez+, 2014)
J/ApJ/829/23 : Stellar flares from Q0-Q17 Kepler LCs (Davenport, 2016)
J/MNRAS/462/2396 : Time-series photometry of IC 348 (Fritzewski+, 2016)
J/AJ/152/114 : Pleiades members with K2 light curves. II. (Rebull+, 2016)
J/A+A/586/A47 : Accretion process in NGC 2264 (Sousa+, 2016)
J/A+A/599/A23 : Accretion-rotation connection in NGC 2264 (Venuti+, 2017)
J/A+A/618/A93 : Gaia DR2 open clusters in Milky Way (Cantat-Gaudin+, 2018)
J/AJ/156/71 : K2 Campaign 2 young disk-bearing in Sco & Oph (Cody+, 2018)
J/AJ/156/84 : APOGEE-2 survey of Orion Complex. II. (Kounkel+, 2018)
J/AJ/157/85 : CIDA Variability Survey of Orion OB1. II. (Briceno+, 2019)
J/AJ/158/122 : Local structure & star formation of the MW (Kounkel+, 2019)
J/ApJ/870/32 : Kinematics young star clusters & associations (Kuhn+, 2019)
J/AJ/158/243 : A search for multiplanet systems with TESS (Pearson, 2019)
J/AJ/158/25 : Automated triage and vetting of TESS candidates (Yu+, 2019)
J/AJ/159/60 : 8695 flares 1228 stars in TESS sectors 1 &2 (Gunther+, 2020)
J/ApJ/919/131 : Oscillating red giants from the TESS QLP (Hon+, 2021)
J/AJ/161/273 : 181 new planet candidates with TESS (Olmschenk+, 2021)
J/ApJ/923/177 : TESS obs. & rotational per of TTauri in Orion (Serna+, 2021)
J/AJ/162/101 : K2 ugri & Ha photometry in the Lagoon Nebula (Venuti+, 2021)
Byte-by-byte Description of file: table7.dat
--------------------------------------------------------------------------------
Bytes Format Units Label Explanations
--------------------------------------------------------------------------------
1- 9 I9 --- TIC TESS input catalog identifier
11- 16 F6.3 mag Tmag [7.62/18.53] TESS magnitude
18- 27 F10.6 deg RAdeg [53/287] TESS Right Ascension (J2000)
29- 38 F10.6 deg DEdeg [-69/37] TESS Declination (J2000)
40- 48 F9.6 d Per [0.18/24.2]? Estimated period
50- 63 A14 --- PRED Prediction, final classification algorithm (1)
65- 72 A8 --- INSPEC Final classification, sample dependent (1)
74- 76 A3 --- flag Additional flags (2)
78- 85 A8 --- sample Sample, testing or training
87- 92 F6.3 --- CLE [0/16.4] contamination level estimator
--------------------------------------------------------------------------------
Note (1): For the "testing" sample PRED refers to the predictions
given by the final classification algorithm while INSPEC refers
to the final classification revised by the visual revision. For
the "training" sample, the PRED field is null (blank), while the
INSPEC field is the assigned class by visual inspection. The
classifications are as follows:
DB = Dipper and/or Bursters classes;
EB = Eclipsing Binaries class;
L = long time variability class;
MP = multiple periods modulated class;
N = noisy class;
NC = No class assigned;
P = periodic light curve;
Pd = periodic light curve with global decrease;
Pda = Modulated class that systematically decreases;
Pi = periodic light curve with global increase;
Pia = Modulated class that systematically increases;
Pn = periodic light curves where noise dominates variability.
Note (2): Additional flags added describing problems in the initial
photometry that affect the criteria of the algorithm. Consider
that these flags are not products of the algorithm, and appeared
only in the visual inspection:
s = Persisting systematic errors dominating the LC;
p = Atypical extreme magnitude values not consistent with the
behavior of the rest of data in the LC;
e = Single transit eclipse in the observational window;
f = Isolated flare in the LC;
r = Very short period star not detected by the LSP;
n = Revised period obtained from the process described in Section 4.3.
--------------------------------------------------------------------------------
History:
From electronic version of the journal
(End) Prepared by [AAS], Coralie Fix [CDS], 09-Apr-2024