Analýza problému parametrizace v hřebenové regresi při řešení multikolinearity.

Abstract

This master's thesis deals with the parametrization of ridge regression when addressing multicollinearity. The aim of the work is to demonstrate how the outputs of regression analysis change with varying input parameters. Through simulation, a method is identified that provides more accurate model estimates. The first part of the thesis describes the theoretical background of the subject matter, while the second part presents its practical application.

Description

Delayed publication

Available after

Subject(s)

Regression models, Regression analysis, Ordinary Least Squares, Ridge regression, Multicollinearity

Citation