Abstract
The success of large-scale data analysis in providing insights and solving complex problems in the data-driven world around us suggests that scientific research can benefit from vast data sets and similar advanced techniques as well. In this dissertation, I demonstrate the validity of this notion by investigating dark matter and redshifts of obscured active galactic nuclei (AGN) using extensive archival X-ray data sets. I present two comprehensive studies of observations peering through the Milky Way's dark matter halo, with the first using ~51 Ms of archival Chandra X-Ray Observatory data and the second using ~41 Ms of archival Swift X-Ray Telescope data, to investigate putative X-ray emission lines from decay of a leading dark matter candidate known as the sterile neutrino, most notably at ~3.5 keV where prior works have detected a feature consistent with sterile neutrino dark matter. The works presented here, together with other similar Milky Way Halo studies that followed the Chandra analysis, provide stringent constraints on sterile neutrino dark matter, which have collectively nearly ruled out the sterile neutrino as a dark matter candidate. I also present a study that employs another archival Chandra data set, which produces a redshift catalog for 121 X-ray-selected obscured AGN with no documented literature values. This broadens the existing census of AGN redshifts, which are critical for piecing together AGN accretion history and its role in galaxy evolution. To compute the redshifts, an existing methodology known as XZ is applied, which can compute obscured AGN redshifts for low-counts X-ray spectra, and an extensive machine learning analysis is performed to improve XZ’s performance on the largely low-redshift data set, for which XZ is not typically well-suited. As a result, in addition to providing the redshift catalog, this work improves the XZ framework and offers a strong blueprint for similar future studies.