Athens University Economics and Business (AUEB)
Statistical monitoring of discrete and nonparametric statistics with exact run length properties
Abstract
dc:descriptionThe current thesis is mainly focused on the design of Univariate nonparametric schemes for monitoring shifts in the process location or scale parameter. Enhanced and robust methodologies are provided which guarantee the chart's optimal and most important reliable statistical performance. In particular, first a short literature review is presented about existing parametric and nonparametric control charts. Additionally, two enhanced distribution-free control charts based on the Sign statistic are introduced, for monitoring shifts in the process location or scale parameter of interest. For each scheme, an extensive numerical analysis is conducted in order to provide to the reader a detailed presentation about each chart's statistical performance. Moreover, we introduce two distribution-free control charts based on the Wilxocon Signed Rank statistic and a simple methodology for the determination of the general distribution of the Wilxocon Signed Rank statistic is provided. Finally, for each proposed scheme, its statistical performance is examined and guidelines to practitioners are given regarding its optimal design. Finally, some concluding remarks and discussion for future work are stated.
Degree
thesis:*- Grantor dc:publisher
- Athens University Economics and Business (AUEB)
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Perdikis, Theodoros
- Περδίκης, Θεόδωρος
Subjects
dc:subject × 10Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/51495