The search for fast optical transients , such as the expected electromagnetic counterparts to binary neutron star mergers , is riddled with false positives ranging from asteroids to stellar flares . While moving objects are readily rejected via image pairs separated by \sim 1 hr , stellar flares represent a challenging foreground that significantly outnumber rapidly-evolving explosions . Identifying stellar sources close to and fainter than the transient detection limit can eliminate these false positives . Here , we present a method to reliably identify stars in deep co-adds of Palomar Transient Factory ( PTF ) imaging . Our machine-learning methodology utilizes the random forest ( RF ) algorithm , which is trained using > 3 \times { 10 } ^ { 6 } sources with Sloan Digital Sky Survey ( SDSS ) spectra . When evaluated on an independent test set , the PTF RF model outperforms the SExtractor star classifier by \sim 4 % . For faint sources ( r ^ { \prime } \geq { 21 } mag ) , which dominate the field population , the PTF RF model produces a \sim 19 % improvement over SExtractor . To avoid false negatives in the PTF transient-candidate stream , we adopt a conservative stellar classification threshold , corresponding to a galaxy misclassification rate = 0.005 . Ultimately , \sim 1.70 \times { 10 } ^ { 8 } objects are included in our PTF point-source catalog , of which only \sim 10 ^ { 6 } are expected to be galaxies . We demonstrate that the PTF RF catalog reveals transients that otherwise would have been missed . To leverage its superior image quality , we additionally create an SDSS point-source catalog , which is also tuned to have a galaxy misclassification rate = 0.005 . These catalogs have been incorporated into the PTF real-time pipelines to automatically reject stellar sources as non-extragalactic transients .