| |
 |
| Paper: |
Automatic Classification of Evolved Objects from the GAIA DR2 and EDR3 Using Machine Learning Tools |
| Volume: |
541, ADASS XXXIII |
| Page: |
197 |
| Authors: |
Silvana G. Navarro; Cynthia A. Martínez-Pinto; Rogelio Hernández; Minia Manteiga; Luis J. Corral |
| DOI: |
10.26624/CLPX5922 |
| Abstract: |
Planetary nebulae (PNe) and symbiotic stars (SY), are both a product
of the evolution of low and medium mass stars. They are not easy to distinguished
with photometric data alone. However the use of some diagnostic diagrams could help
to distinguish these objects between them and from other type of evolved objects like
Red Giants, Mira, cataclysmic variables, etc. We present the results of the automatic
classification based on GAIA photometry data from releases DR2 and EDR3. The
classification was made using different algorithms and the results compared in basis
of their accuracy. The training catalogue was constructed using the GAIA parameters
(Gmag, BP mag and RP mag) which were complemented with J, H and K magnitudes from the 2MASS catalogue and some b−v colors when they were available from
SIMBAD database. We present the results concerning the accuracy obtained and the
better combination of parameters to achieve the best effectiveness. It was found that
the b−v color, used frequently to separate NPs from SY, can be replaced by GAIA
colors: Gmag−BPmag or BPmag−RPmag with advantage over b−v in some diagnostic diagrams, due to the wider availability of GAIA’s indices. The inclusion of red
giants on the database show the inherent difficulties to successful separate them due to
the wide range of colors they can present. The access to the GAIA spectra will allow
us to separate these objects from the NPs and SS (work in progress). |
|
|
|
|
|
