Record Details
Parameter estimation of Gaussian hierarchical model using Gibbs sampling
ScholarsArchive at Oregon State University
| Field | Value |
|---|---|
| Title | Parameter estimation of Gaussian hierarchical model using Gibbs sampling |
| Names |
Mbuthia, Juliana
(creator) Thinh, Nguyen (advisor) |
| Date Issued | 2014-06-04 (iso8601) |
| Note | Graduation date: 2015 |
| Abstract | Gibbs sampling method is an important tool used in parameter estimation for many probabilistic models. Specifically, for many scenarios, it is difficult to generate high-dimensional data samples from its joint distribution. The Gibbs sampling provides a way to draw high-dimensional data via the conditional distributions which are typically easier to sample. In this thesis, we study a simple generative model called Hierarchical Gaussian and an efficient method for computing its parameters using Gibbs sampling. In particular, we show that the Hierarchical Gaussian model admits closed form conditional distributions such that Gibbs sampling can be used effectively to draw the samples from the joint distribution, and perform parameter estimation. |
| Genre | Thesis/Dissertation |
| Topic | Parameter estimation |
| Identifier | http://hdl.handle.net/1957/51386 |
