Astronomical surveys of celestial sources produce streams of noisy time series measuring flux versus time ("light curves"). Unlike in many other physical domains, however, large (and source-specific) temporal gaps in data arise naturally due to intranight cadence choices as well as diurnal and seasonal constraints. With nightly observations of millions of variable stars and transients from upcoming surveys, efficient and accurate discovery and classification techniques on noisy, irregularly sampled data must be employed with minimal human-in-the-loop involvement. Machine learning for inference tasks on such data traditionally requires the laborious hand-coding of domain-specific numerical summaries of raw data ("features"). Here we present a novel unsupervised autoencoding recurrent neural network (RNN) that makes explicit use of sampling times and known heteroskedastic noise properties. When trained on optical variable star catalogs, this network produces supervised classification models that rival other best-in-class approaches. We find that autoencoded features learned on one time-domain survey perform nearly as well when applied to another survey. These networks can continue to learn from new unlabeled observations and may be used in other unsupervised tasks such as forecasting and anomaly detection.

The Canarias InfraRed Camera Experiment (CIRCE) is a near-infrared (1-2.5 micron) imager, polarimeter and low-resolution spectrograph operating as a visitor instrument for the Gran Telescopio Canarias 10.4-meter telescope. It was designed and built largely by graduate students and postdocs, with help from the UF astronomy engineering group, and is funded by the University of Florida and the U.S. National Science Foundation. CIRCE is intended to help fill the gap in near-infrared capabilities prior to the arrival of EMIR to the GTC, and will also provide the following scientific capabilities to compliment EMIR after its arrival: high-resolution imaging, narrowband imaging, high-time-resolution photometry, imaging polarimetry, low resolution spectroscopy. In this paper, we review the design, fabrication, integration, lab testing, and on-sky performance results for CIRCE. These include a novel approach to the opto-mechanical design, fabrication, and alignment.

We present precise V and Ks-band transit photometry for the planetary host star OGLE-TR-113. Using the Ks-band photometry, we confirm the dwarf nature of OGLE-TR-113, and obtain new estimates for its effective temperature, distance and reddening. We employ the V-band photometry to obtain planetary and orbit parameters from the transit fit, a= (0.0232 \pm 0.0038) AU, orbital period P= (1.4324752 \pm 0.0000015) days, i= 86.7 - 90, R_p= (1.09 \pm 0.09) R_J. These values are in excellent agreement with previous works. Assuming a mass M_p= (1.32 \pm 0.19) M_J for the planet we obtain its mean density \rho= (1.26 \pm 0.50) g cm^{-3}, also in agreement with previous works. The transit observed in the Ks-band has a larger scatter and we find its amplitude to be consistent with that in the V-band. In this way, we find an independent confirmation of the planetary nature of OGLE-TR-113b.