A Next Generation Analysis Pipeline for 21cm Epoch of Reionization StudiesKerrigan, JoshuacreatorPober, JonathanAdvisorTucker, GregoryReaderDell'Antonio, IanReaderBrown University. Department of Physicssponsor201913, 130 p.born digitalThesis (Ph. D.)--Brown University, 2019thesesThe Epoch of Reionization (EoR) is a period during the history of the Universe that as of today has been largely unconstrained by cosmological measurements. The most promising probe of reionization, the 21cm emission of neutral hydrogen, is unfortunately difficult to observe. Measurements of the 21cm reionization signal are performed using low-frequency wide-field radio interferometers. The difficulty in detecting the weak 21cm signal is further compounded by instrumental and astrophysical limitations. In an effort to overcome these issues, I present four key results. First I look at applying a hybrid foreground avoidance strategy for retaining sensitivity in the presence of Galactic and extragalactic foregrounds. I then introduce a novel method for identifying radio frequency interference (RFI) in interferometric visibilities using a Deep Fully Convolutional Neural Network (DFCN). Following this, I perform sky calibration on preliminary Hydrogen Epoch of Reionization Array (HERA) observations using the Fast Holographic Deconvolution (FHD) interferometry simulation software which provides an independent validation of the default HERA processing and calibration pipeline. Finally, I demonstrate a Convolutional Neural Network (CNN) for 21cm parameter estimation and uncertainty propagation on simulated 21cm cosmological volumes.Machine LearningCosmologyRadio astronomyEpoch of reionization21cmEnglishRPB2019060310.26300/gd9x-ms58In CopyrightCollection is open for research.dissertations