|
|
Comparison of exact and asymptotic tests
Rusnáková, Viktória ; Hlubinka, Daniel (advisor) ; Došlá, Šárka (referee)
Nazev prace: Porovnani pfesnych a asymptotickych testu Autor: Viktoria Rusnakova Katedra: Katcdra pravdepodobnosti a matematicke statistiky Vedouci bakalafske prace: Doc. RNDr. Daniel Hlubinka, Ph.D. e-mail vedouciho: hlubinka(Q)karlin.mff.cuni.cz Abstrakt: Predlozena praca sa zaobera testami o parametroch binomickeho. Poiaso novho a cxponencialneho rozdclenia. Ciol'om prace je porovnat! presne znahodneiie testy a testy asymptoticke, zalozcne na centralnej limitnej veto a zalozcne na metode maximalnej vierohodnosti. V prvej, teoretickej casti su uvedene potrebne vety a tvrdcnia, na zaklado ktorych su zostrojenc priklady pro dane rozdclenia. Druha, prakticka cast' je venovana pocitacovym simulaciarn obojstrannych testov na hlacline vyznamnosti 5 %. Vysledky siraulacii su uvedene na prilozenom CD. Klfcova slova: kriticka funkcia, znahodneny test, asymptoticky test Title: Comparison of exact and asymptotic tests Author: Viktoria Rusnakova Department: Department of Probability and Mathematical Statistics Supervisor: Doc. RNDr. Daniel Hlubinka, Ph.D. Supervisor's e-mail address: hlubinkaykarhn.mff.cuin.cz Abstract: The presented work deals with tests about parameters of binomial, Poisson and exponencial distribution. The aim of this work is to compare exact randomized tests and asymptotic tests based on central...
|
|
|
Sliced Inverse Regression
Rusnáková, Viktória ; Hlávka, Zdeněk (advisor) ; Lachout, Petr (referee)
The presented work deals with Sliced inverse regression method for dimension reduction of explanatory variable in case of multivariate data sets. The aim of this work is to present this method, its characteristics and assumptions and ways of determining the new dimension of explanatory variable. This is done in the first theoretical part. The second practical part is dedicated to application of SIR method on simulated and real data. All simulations and calculations were done in programme R and are enclosed on CD.
|
|
|
Sliced Inverse Regression
Rusnáková, Viktória ; Hlávka, Zdeněk (advisor) ; Lachout, Petr (referee)
The presented work deals with Sliced inverse regression method for dimension reduction of explanatory variable in case of multivariate data sets. The aim of this work is to present this method, its characteristics and assumptions and ways of determining the new dimension of explanatory variable. This is done in the first theoretical part. The second practical part is dedicated to application of SIR method on simulated and real data. All simulations and calculations were done in programme R and are enclosed on CD.
|
|
|
Comparison of exact and asymptotic tests
Rusnáková, Viktória ; Došlá, Šárka (referee) ; Hlubinka, Daniel (advisor)
Nazev prace: Porovnani pfesnych a asymptotickych testu Autor: Viktoria Rusnakova Katedra: Katcdra pravdepodobnosti a matematicke statistiky Vedouci bakalafske prace: Doc. RNDr. Daniel Hlubinka, Ph.D. e-mail vedouciho: hlubinka(Q)karlin.mff.cuni.cz Abstrakt: Predlozena praca sa zaobera testami o parametroch binomickeho. Poiaso novho a cxponencialneho rozdclenia. Ciol'om prace je porovnat! presne znahodneiie testy a testy asymptoticke, zalozcne na centralnej limitnej veto a zalozcne na metode maximalnej vierohodnosti. V prvej, teoretickej casti su uvedene potrebne vety a tvrdcnia, na zaklado ktorych su zostrojenc priklady pro dane rozdclenia. Druha, prakticka cast' je venovana pocitacovym simulaciarn obojstrannych testov na hlacline vyznamnosti 5 %. Vysledky siraulacii su uvedene na prilozenom CD. Klfcova slova: kriticka funkcia, znahodneny test, asymptoticky test Title: Comparison of exact and asymptotic tests Author: Viktoria Rusnakova Department: Department of Probability and Mathematical Statistics Supervisor: Doc. RNDr. Daniel Hlubinka, Ph.D. Supervisor's e-mail address: hlubinkaykarhn.mff.cuin.cz Abstract: The presented work deals with tests about parameters of binomial, Poisson and exponencial distribution. The aim of this work is to compare exact randomized tests and asymptotic tests based on central...
|
guest ::
RSS feed.