Your browser does not allow JavaScript!
JavaScript is necessary for the proper functioning of this website. Please enable JavaScript or use a modern browser.
Repository of the University of Ljubljana
Open Science Slovenia
Open Science
DiKUL
slv
|
eng
Search
Browse
New in RUL
About RUL
In numbers
Help
Sign in
Details
Iterativne numerične metode v posplošenih linearnih modelih : delo diplomskega seminarja
ID
Mandić, Mitja
(
Author
),
ID
Smrekar, Jaka
(
Mentor
)
More about this mentor...
PDF - Presentation file,
Download
(952,96 KB)
MD5: D82334AAB0ECB94899FD111F1CA22E8D
Image galllery
Abstract
V nalogo smo se spustili z namenom razumeti postopek ocenjevanja parametrov v posplošenih linearnih modelih. Za uvod si postavimo teoretične temelje z eksponentno družino in izpeljemo nekaj lastnosti. Nato definiramo posplošene linearne modele in si ogledamo nekaj najpomembenjših primerov ter predstavimo metode za ocenjevanje parametrov, s poudarkom na metodi največjega verjetja. Izpeljemo enačbe verjetja za logistični model, rezultate nato komentiramo v luči eksponentne družine in jih posplošimo na vse porazdelitve, ki ji pripadajo. Izpeljemo tudi enačbe verjetja v probit modelu in opazimo prednosti uporabe kanoničnih povezovalnih funkcij. V drugem delu naloge se posvetimo numeričnim metodam. Izpeljemo Newtonovo metodo in komentiramo težave, ki lahko nastopijo z njeno uporabo. Izpeljemo tudi Fisherjevo zbirno metodo in dokažemo, da se ob uporabi modela s kanonično povezovalno funkcijo ujema z Newtonovo metodo. Izpeljano teorijo v zadnjem delu povežemo v praktičnem primeru. Primerjamo rezultate dobljene s probit in logističnim modelom in komentiramo morebitne razlike.
Language:
Slovenian
Keywords:
eksponentna družina
,
kanonični parameter
,
cenilka največjega verjetja
,
logistični model
,
Fisherjeva zbirna metoda
Work type:
Final seminar paper
Typology:
2.11 - Undergraduate Thesis
Organization:
FMF - Faculty of Mathematics and Physics
Year:
2021
PID:
20.500.12556/RUL-130275
UDC:
519.2
COBISS.SI-ID:
76276227
Publication date in RUL:
12.09.2021
Views:
1133
Downloads:
110
Metadata:
Cite this work
Plain text
BibTeX
EndNote XML
EndNote/Refer
RIS
ABNT
ACM Ref
AMA
APA
Chicago 17th Author-Date
Harvard
IEEE
ISO 690
MLA
Vancouver
:
MANDIĆ, Mitja, 2021,
Iterativne numerične metode v posplošenih linearnih modelih : delo diplomskega seminarja
[online]. Bachelor’s thesis. [Accessed 17 May 2025]. Retrieved from: https://repozitorij.uni-lj.si/IzpisGradiva.php?lang=eng&id=130275
Copy citation
Share:
Secondary language
Language:
English
Title:
Iterative numerical methods in generalized linear models
Abstract:
We have conducted the following research in order to understand the process of parameter estimation in generalized linear models. In the beginning we lay theoretical foundations with exponential family and derive some of it's properties. Then we define generalized linear models, inspect some more important cases and define multiple methods for parameter estimation, taking a closer look at the maximum likelihood method. We go on to derive maximum likelihood equations in the logistic model and generalize the result for the exponential family. As an alternative, we derive the same equations also for the probit model and comment on the advantages of using canonical link functions. The second part focuses on numerical methods. We derive the Newton method and comment on its possible issues. We also define the Fisher's scoring algorithm and prove the equivalence of the methods for canonical distribution models. Theory is then put to work in the last part of the research. We compare probit and logit models and comment on the differences between the two.
Keywords:
exponential family
,
canonical parameter
,
maximum likelihood estimator
,
logistic model
,
Fisher's scoring algorithm
Similar documents
Similar works from RUL:
Searching for similar works...
Similar works from other Slovenian collections:
Back