Μέθοδοι bootstrap στη γραμμική παλινδρόμηση
Bootstrap methods in linear regression
View/
Keywords
Bootstrap ; Linear regressionAbstract
Linear regression constitutes one of the most important statistical methods for the simultaneous analysis of two or more variables, with the aim of predicting one of them based on the known value(s) of one or more related variables. The application of the linear regression requires the fulfilment of specific assumptions, such as normality, homoscedasticity, and independence of the error terms. In practice, these assumptions are frequently violated, which can lead to unreliable conclusions. Bootstrap methods provide robust non-parametric alternatives, as they allow for the estimation of standard errors and confidence intervals based solely on the observed data. The present thesis presents and compares various bootstrap approaches in linear regression, including the parametric bootstrap, the residual bootstrap, the paired bootstrap, and the wild bootstrap, through simulation studies and applications to real datasets.
Postgraduate Studies Programme
Εφαρμοσμένη ΣτατιστικήDepartment
Σχολή Χρηματοοικονομικής και Στατιστικής. Τμήμα Στατιστικής και Ασφαλιστικής ΕπιστήμηςNumber of pages
60Language
GreekCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα