The GALAH Survey: A New Sample of Extremely Metal-poor Stars Using a Machine-learning Classification Algorithm

Extremely Metal-Poor (EMP) stars provide a valuable probe of early chemical enrichment in the Milky Way. Here we leverage large sample $\sim600,000$ high-resolution stellar spectra from GALAH survey plus machine learning algorithm to find 54 candidates with estimated [Fe/H]~$\leq$~-3.0, 6 which have [Fe/H]~$\leq$~-3.5. Our includes $\sim 20 \%$ main sequence EMP candidates, unusually high for \emp surveys. We magnitude-limited metallicity distribution function our is consistent previous work that used more complex selection criteria. The method present has significant potential application next generation massive spectroscopic surveys, will expand available data well into millions stars.