Beijesa empīriskās ticamības metode
Loading...
Date
Authors
Advisor
Journal Title
Journal ISSN
Volume Title
Publisher
Latvijas Universitāte
Language
lav
Abstract
Darbs veltīts Beijesa empīriskās ticamības metodei. Izmantojot Hamiltona Montekarlo algoritmu, ilustrēti pieejas lietojumi vidējās vērtības, lineārās un loģistiskās regresijas koeficientu, kā arī korelācijas koeficienta novērtēšanai. Analizētas arī Beijesa empīriskās ticamības metodes statistiskās īpašības, salīdzinot pārklājuma precizitāti un empīrisko jaudu ar parametrisko t-testu un empīriskās ticamības pieeju.
The master’s thesis is devoted to the Bayesian empirical likelihood. Using the Hamiltonian Monte Carlo algorithm, applications of the approach for estimating the mean, linear and logistic regression coefficients, and the correlation coefficient are illustrated. Additionally, the statistical properties of the Bayesian empirical likelihood are analyzed, comparing the coverage accuracy and empircal power with the parametric t-test and the empirical likelihood approach.
The master’s thesis is devoted to the Bayesian empirical likelihood. Using the Hamiltonian Monte Carlo algorithm, applications of the approach for estimating the mean, linear and logistic regression coefficients, and the correlation coefficient are illustrated. Additionally, the statistical properties of the Bayesian empirical likelihood are analyzed, comparing the coverage accuracy and empircal power with the parametric t-test and the empirical likelihood approach.