Thesis (Selection of subject)Thesis (Selection of subject)(version: 385)
Thesis details
   Login via CAS
Optimální regresní dizajn
Thesis title in Czech: Optimální regresní dizajn
Thesis title in English: Optimal regression design
Key words: regresní model|dizajn experimentu|optimální dizajn
English key words: regression moidel|experimental design|optimal design
Academic year of topic announcement: 2025/2026
Thesis type: diploma thesis
Thesis language:
Department: Department of Probability and Mathematical Statistics (32-KPMS)
Supervisor: doc. RNDr. Matúš Maciak, Ph.D.
Author:
Guidelines
V klasickom lineárnom regresnom modeli a odhadovaní neznámych regresných parametrov pomocou metódy najmenších štvorcov
vyvstáva pri plánovaní experimentu otázka týkajúca sa vhodnosti a efektívnosti navrhnovaného dizajnu.
Existujú ale rôzne kvalitatívne a kvantitatívne kritéria ako posudzovať vhodnosť a efektívnosť konkretného návrhu experimentu.
Väčšina týchto kritérii je založená na určitej optimalizácii variančnej-kovariančnej matici odhadovaných parametrov (pomocou metódy najmenších štvorcov).

Cieľom tejto práce je oboznámiť sa s tzv. štandardizovanými kritériami optimality regresného dizajnu.
Autor jednak sumarizuje rôzne postupy existujúce v odbornej literatúre, porovná ich základné teoretické vlastnosti (výhody a nevýhody)
a pomocou simulácii vyšetri fungovanie na datach.

References
[1] Dette, H. (1997). Designing Experiments with Respect to 'Standardized' Optimality Criteria. Journal of the Royal Statistical Society, Series B., Vol.59, No.1, 97--110.

[2] Kiefer, J. (1971). On the role of symmetry and approximation in exact design optimality. Statistical Decision Theory and Related Topics (Eds S. Gupta and J. Yackel), 109--118. New York Academic Press.

[3] Pázman, A. (1986). Foundations of optimum experimental design. Dordrecht: D. Reidel Publishing Company.

[4] Pukelsheim, F. (1993). Optimal Design of Experiments. John Wiley & Sons, New York.
Preliminary scope of work in English
In the design of experiments, optimal experimental designs are designs that are optimal with respect to some well-defined (explicit) statistical criterion.
There are of course many different optimality criteria defined in the literature. However, the fundamental assumption in the optimal design theory is that the underlying statistical
model is known. This thesis will focus on the situation where the underlying model is an ordinary (multiple) linear regression model and the vector of the unknown regression parameters
is estimated in terms of the least squares minimization.
 
Charles University | Information system of Charles University | http://www.cuni.cz/UKEN-329.html