Ticamības intervāli nošķeltajai vidējai vērtībai
Loading...
Date
Authors
Advisor
Journal Title
Journal ISSN
Volume Title
Publisher
Latvijas Universitāte
Language
lav
Abstract
Šinī darbā tiek apskatīts nošķeltās vidējās vērtības ticamības intervāla aprēķins pēc divām metodēm - izmantojot asimptotisko normalitāti vai izmantojot empīriskās ticamības attiecību. Darba teorētiskajā daļā tiek aprakstītas abas šīs metodes un to priekšnosacījumi. Praktiskajā daļā veicot simulācijas tiek pārbaudīta pārklājuma precizitāte katrai no metodēm pie izlasēm no dažādām sadalījuma funkcijām. Darbam pievienots programmēšanas valodā R izveidotais kods.
This work is devoted to trimmed mean confidence interval calculations based on two methods - either using asymptotic normality of trimmed mean or using empirical likelihood ratio. In theoretical part of the work are described both of these methods and their prerequisites. In practical part of the work coverage accuracy of each method is examined using simulations and generating samples from an array of different distribution functions. Attached is the R code that was developed as part of this work.
This work is devoted to trimmed mean confidence interval calculations based on two methods - either using asymptotic normality of trimmed mean or using empirical likelihood ratio. In theoretical part of the work are described both of these methods and their prerequisites. In practical part of the work coverage accuracy of each method is examined using simulations and generating samples from an array of different distribution functions. Attached is the R code that was developed as part of this work.