AbstractNarrow-band photometric surveys, such as the Javalambre Photometric Local Universe Survey (J-PLUS), provide not only a means of pre-selection for high-resolution follow-up, but open a new era of precision photometric stellar parameter determination. Using a family of machine learning algorithms known as Artificial Neural Networks (ANNs), we have obtained photometric estimates of effective temperature (T_eff) and metallicity ([Fe/H]) across a wide parameter range of temperature and metallicity (4000 < T_eff <7000 K; −3.5 <[Fe/H]<0.0) for a number of stars in the J-PLUS Early Data Release. With this methodology, we expect to increase the number of known Carbon-enhanced Metal-poor (CEMP; [C/Fe]>+0.7) stars by several orders of magnitude, as well as constrain the metallicity distribution function of the Milky Way Halo system.