VizieR Online Data Catalog: M-type stars in LAMOST DR5 unclassified spectra (Guo+, 2019)

LAMOST is a 4-m reflecting Schmidt telescope with a large field of view (FoV) of 5 degrees in diameter. It has 4000 fibres mounted on its focal plane and 16 spectrographs with 32 CCD cameras, so that it can simultaneously observe up to 4000 objects (Cui et al. 2012RAA....12.1197C). The raw CCD data are reduced and analyzed by the LAMOST data pipelines, which consist of a 2D pipeline and a 1D pipeline. The primary functions of the 2D pipeline include bias calibration, flat-field correction, spectral extraction, sky subtraction, wavelength calibration, flux calibration, and sub-exposures combination. The calibrated spectra from the 2D pipeline are then fed to the 1D pipeline which performs spectral classification and parameter determination based on template matching and chi-square criteria (Luo et al. 2015RAA....15.1095L, Cat. V/146).

By July 2017, LAMOST had completed its five-year regular survey. The LAMOST DR5 V1 includes 9017844 spectra of stars, galaxies, quasars, and unrecognized objects. These spectra cover the wavelength range from 3690 to 9100Å with a resolution of R~1800 at the wavelength 5500Å.

Among the 9 million spectra in LAMOST DR5 V1, 642178 unrecognized spectra were labelled 'UNKNOWN'. During the classification process of the 1D pipeline, a spectrum is classified as 'UNKNOWN' if the confidence of the classification result is lower than a given threshold value. In this study we build a multilayer Pseudo-Inverse Learning (ML-PIL) to fulfil the hash-learning process, so as to search M-type stars in the 'UNKNOWN' spectra of LAMOST DR5 V1.

(1 data file).