Identifying X-ray binary (XRB) candidates in nearby galaxies requires distinguishing them from possible contaminants including foreground stars and background active galactic nuclei. This work investigates the use of supervised machine learning algorithms to identify high-probability X-ray binary candidates. Using a catalogue of 943 Chandra X-ray sources in the Andromeda galaxy, we trained and tested several classification algorithms using the X-ray properties of 163 sources with previously known types. Amongst the algorithms tested, we find that random forest classifiers give the best performance and work better in a binary classification (XRB/non-XRB) context compared to the use of multiple classes. Evaluating our method by comparing with classifications from visible-light and hard X-ray observations as part of the Panchromatic Hubble Andromeda Treasury, we find compatibility at the 90% level, although we caution that the number of sources in common is rather small. The estimated probability that an object is an X-ray binary agrees well between the random forest binary and multiclass approaches and we find that the classifications with the highest confidence are in the X-ray binary class. The most discriminating X-ray bands for classification are the 1.7-2.8, 0.5-1.0, 2.0-4.0, and 2.0-7.0 keV photon flux ratios. Of the 780 unclassified sources in the Andromeda catalogue, we identify 16 new high-probability X-ray binary candidates and tabulate their properties for follow-up.

We present an unprecedented, deep study of the primordial low-mass X-ray binary population in an isolated, lower-metallicity environment. We perform followup observations of previously-identified X-ray binary candidates in the Sculptor Dwarf Galaxy by combining a second Chandra observation with Spitzer and Gemini photometry, as well as Gemini spectroscopy of selected targets. Of the original nine bright X-ray sources identified, we are able to classify all but one as quasars, active galactic nuclei, or background galaxies. We further discover four new X-ray sources in the second-epoch Chandra observation. Three of these new sources are background sources and one is a foreground flaring star. We have found that Sculptor is effectively devoid of X-ray sources above a few 1e34 erg/s. If Sculptor is able to retain primordial binaries at a similar rate to globular clusters, this implies that bright X-ray binaries observed in globular clusters in the present epoch are all formed dynamically.

The observed relation between the X-ray and radio properties of low-luminosity accreting black holes has enabled the identification of multiple candidate black hole X-ray binaries (BHXBs) in globular clusters. Here we report an identification of the radio source VLA J213002.08+120904 (aka M15 S2), recently reported in Kirsten et al. 2014, as a BHXB candidate. They showed that the parallax of this flat-spectrum variable radio source indicates a 2.2$^{+0.5}_{-0.3}$ kpc distance, which identifies it as lying in the foreground of the globular cluster M15. We determine the radio characteristics of this source, and place a deep limit on the X-ray luminosity of $\sim4\times10^{29}$ erg s$^{-1}$. Furthermore, we astrometrically identify a faint red stellar counterpart in archival Hubble images, with colors consistent with a foreground star; at 2.2 kpc its inferred mass is 0.1-0.2 $M_{\odot}$. We rule out that this object is a pulsar, neutron star X-ray binary, cataclysmic variable, or planetary nebula, concluding that VLA J213002.08+120904 is the first accreting black hole X-ray binary candidate discovered in quiescence outside a globular cluster. Given the relatively small area over which parallax studies of radio sources have been performed, this discovery suggests a much larger population of quiescent BHXBs in our Galaxy, $2.6\times10^4-1.7\times10^8$ BHXBs at $3\sigma$ confidence, than has been previously estimated ($\sim10^2-10^4$) through population synthesis.

We present near-simultaneous Chandra/HST observations of the very faint ($L_{x} < 10^{36}$ erg s$^{-1}$) X-ray transient source M15 X-3, as well as unpublished archival Chandra observations of M15 X-3. The Chandra observations constrain the luminosity of M15 X-3 to be $< 10^{34}$ erg s$^{-1}$ in all observed epochs. The X-ray spectrum shows evidence of curvature, and prefers a fit to a broken power-law with break energy $E_{\rm break} = 2.7^{+0.4}_{-0.6}$ keV, and power law indices of $\Gamma_{1} = 1.3^{+0.1}_{-0.2}$ and $\Gamma_{2} = 1.9^{+0.2}_{-0.2}$ over a single power law. We fit our new F438W ($B$), F606W (broad $V$), and F814W ($I$) HST data on the blue optical counterpart with a model for an accretion disk and a metal-poor main sequence star. From this fit, we determine the companion to be consistent with a main sequence star of mass $0.440^{+0.035}_{-0.060}$ $M_{\odot}$ in a $\sim$4-hour orbit. X-ray irradiation of the companion is likely to be a factor in the optical emission from the system, which permits the companion to be smaller than calculated above, but larger than $0.15$ $M_{\odot}$ at the $3\sigma$ confidence level. M15 X-3 seems to be inconsistent with all suggested hypotheses explaining very faint transient behavior, except for magnetospherically inhibited accretion.