1611.07526
The weirdest SDSS galaxies: results from an outlier detection algorithm
Baron, Poznanski
Present an outlier detection algorithm, based on an unsupervised Random Forest. Test the algorithm on more than 2M spectra from SDSS and examine the 400 galaxies with the highest outlier score. Find objects which have extreme emission line ratios and abnormally strong absorption lines, objects with unusual continua, including extremely reddened galaxies. Find galaxy-galaxy gravitational lenses, double-peaked emission line galaxies, and close galaxy pairs. Find galaxies with high ionization lines, galaxies which host See, and galaxies with unusual gas kinematics. Only a fraction of the outliers found were reported by previous studies that used specific and tailored algorithms to find a single class of unusual objects. The algorithm is general and detects all of these classes, and many more, regardless of what makes them peculiar. It can be executed on imaging, time-serious, and other spectroscopic data, operates well with 1000s of features, is not sensitive to missing values, and is easily parallelisable.
1611.07578
2dFLenS and KiDS: determining source redshift distributions with cross-correlations
Johnson, Blake, Amon, Erben, et al
Develop a statistical estimator to infer the z probability distribution of a photometric sample of galaxies from its angular cross-correlation in z bins with an overlapping spectroscopic sample. This estimator is a minimum variance weighted quadratic function of the data: a quadratic estimator. This extends and modifies the methodology presented by McQuinn&White (2013). The derived source z distribution is degenerate with the source galaxy bias, which must be constrained via additional assumptions. Apply this estimator to constrain source galaxy z distributions in the KiDS imaging survey through cross-correlation with the spectroscopic 2-degree Field Lensing Survey, presenting results first as a binned step-wise distribution in the range z<0.8, and then building a continuous distribution using a Gaussian process model. Demonstrate the robustness of the methodology using mock catalogues constructed from N-body sims, and comparisons with other techniques for inferring the z distribution.
No comments:
Post a Comment