Files in this item



application/pdf3392011.pdf (7MB)Restricted to U of Illinois
(no description provided)PDF


Title:Precision Cosmological Parameter Estimation
Author(s):Fendt, William Ashton, Jr
Doctoral Committee Chair(s):Brunner, Robert J.
Department / Program:Physics
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Physics, Astronomy and Astrophysics
Abstract:Our idea is to shift the computationally intensive pieces of the parameter estimation framework to a parallel training step. We then provide a machine learning code that uses this training set to learn the relationship between the underlying cosmological parameters and the function we wish to compute. This code is very accurate and simple to evaluate. It can provide incredible speed-ups of parameter estimation codes. For some applications this provides the convenience of obtaining results faster, while in other cases this allows the use of codes that would be impossible to apply in the brute force setting. In this thesis we provide several examples where our method allows more accurate computation of functions important for data analysis than is currently possible. As the techniques developed in this work are very general, there are no doubt a wide array of applications both inside and outside of cosmology. We have already seen this interest as other scientists have presented ideas for using our algorithm to improve their computational work, indicating its importance as modern experiments push forward. In fact, our algorithm will play an important role in the parameter analysis of Planck, the next generation CMB space mission.
Issue Date:2009
Description:148 p.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2009.
Other Identifier(s):(MiAaPQ)AAI3392011
Date Available in IDEALS:2015-09-25
Date Deposited:2009

This item appears in the following Collection(s)

Item Statistics