Bayesian model selection

Phuong Thi Thanh Truong

Diplomová práce

Anotace:

Model selection is an important problem in many branches including statistical analysis. In this thesis, we use the Bayesian model-selection to find the model that is the most suitable for the given data. We begin with Bayesian inference and theoretical background of Bayesian model selection. We derive the posterior distribution and marginal likelihood for linear regression models. By using Bayesian inference in regression models, we can identify the best model from generated datasets.

Abstract:

Model selection is an important problem in many branches including statistical analysis. In this thesis, we use the Bayesian model-selection to find the model that is the most suitable for the given data. We begin with Bayesian inference and theoretical background of Bayesian model selection. We derive the posterior distribution and marginal likelihood for linear regression models. By using Bayesian inference in regression models, we can identify the best model from generated datasets.

Klíčová slova

Bayesian statisstics, linear regression, model selection.

Keywords

Bayesian statisstics, linear regression, model selection.

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 30. 4. 2019

Obhajoba závěrečné práce

• Obhajoba proběhla 29. 5. 2019
• Vedoucí: Jan Kracík
• Oponent: Thanh Tien Thach

VŠB - Technická univerzita Ostrava

Fakulta elektrotechniky a informatiky

Magisterský studijní program / obor:
Informační a komunikační technologie / Výpočetní matematika