Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
GPAbin to unify principal component analysis biplots from multiple imputations
Thesis (MCom)--Stellenbosch University, 2025.
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | en_ZA |
| Published: |
Stellenbosch : Stellenbosch University
2025
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867614075249229825 |
|---|---|
| access_status_str | Open Access |
| author | Ramaisa, Mokgeseng |
| author2 | Nienkemper-Swanepoel, J. |
| author_browse | Nienkemper-Swanepoel, J. Ramaisa, Mokgeseng |
| author_facet | Nienkemper-Swanepoel, J. Ramaisa, Mokgeseng |
| author_sort | Ramaisa, Mokgeseng |
| collection | Thesis |
| dc_rights_str_mv | Stellenbosch University |
| description |
Thesis (MCom)--Stellenbosch University, 2025. |
| format | Thesis |
| id | oai:scholar.sun.ac.za:10019.1/132480 |
| institution | Stellenbosch University (South Africa) |
| language | en_ZA |
| last_indexed | 2026-06-10T12:46:15.146Z |
| license_str | Other — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from SUNScholar — Stellenbosch University Repository |
| publishDate | 2025 |
| publishDateRange | 2025 |
| publishDateSort | 2025 |
| publisher | Stellenbosch : Stellenbosch University |
| publisherStr | Stellenbosch : Stellenbosch University |
| record_format | dspace |
| source_str | SUNScholar — Stellenbosch University Repository |
| spelling | oai:scholar.sun.ac.za:10019.1/132480 GPAbin to unify principal component analysis biplots from multiple imputations Ramaisa, Mokgeseng Nienkemper-Swanepoel, J. Stellenbosch University. Faculty of Economic and Management Sciences. Dept. of Statistical & Actuarial Science. Multivariate analysis -- Graphic methods Correspondence analysis (Statistics) Biplots Graphical modeling (Statistics) Multiple comparisons (Statistics) UCTD Thesis (MCom)--Stellenbosch University, 2025. Ramaisa, M. 2025. GPAbin to unify principal component analysis biplots from multiple imputations. Unpublished masters thesis. Stellenbosch: Stellenbosch University [online]. Available: https://scholar.sun.ac.za/items/d0c890f4-f178-403e-9e1d-1dfca4ba441a ENGLISH SUMMARY: Missing data in real world applications are frequently encountered by data practitioners. The strategy to handle missing data is often deletion which results in loss of information and biased results in analyses. It is important to investigate the underlying reason for missingness to decide on an appropriate handling strategy to reduce bias. Multiple imputation is a superior technique to handle missing values, in which multiple possible values for missing data are imputed, resulting in multiple completed data sets. A practitioner may be interested in exploratory data analysis through visualisation, however interpreting and analysing multiple visualisations of completed data sets may lead to subjective bias and may be time intensive. GPAbin has been developed to unify multiple completed multivariate categorical data visualisations, specifically multiple correspondence analysis biplots. This has been achieved using generalised orthogonal Procrustes analysis and Rubin's rules to result in a single unbiased visualisation for practitioners. The extension of the GPAbin methodology to continuous completed data sets is presented in this project. This is achieved by utilising the extension of the classic Gabriel principal component analysis biplot as a starting point. A simulation study is presented to further understand the performance of the newly developed methodology. AFRIKAANSE OPSOMMING: Ontbrekende data work gereeld deur data-praktisyns teegekom in toegepaste probleme. Die strategie om ontbrekende data te hanteer is dikwels weglating, wat lei tot verlies van inligting en sydigheid in ontledings. Dit is belangrik om die onderliggende rede vir ontbreking te ondersoek om te besluit op 'n toepaslike analise strategie om sydigheid te verminder. Meervoudige imputasie, waarin verskeie moontlike waardes vir ontbrekende data imputasie word, is 'n meer gepaste tegniek om ontbrekende waardes te hanteer wat lei tot meervoudige voltooide datastelle. 'n Praktisyn kan belangstel in verkennende data-analise deur middel van visualisering, maar die interpretasie en ontleding van verskeie visualiserings van voltooide datastelle mag lei tot subjektiewe sydigheid en kan tydsintensief wees. GPAbin is ontwikkel om meervoudige voltooide meerveranderlike kategoriese datavisualiserings, spesifiek meervoudige ooreenkomsanalise bi-stippings te verenig. Dit is bereik deur veralgemeende ortogonale Procrustes-analise en Rubin se reels te gebruik om 'n enkele onsydige visualisering vir praktisyns tot gevolg te he. Die uitbreiding van die GPAbin metodologie tot kontinue voltooide datastelle word in hierdie projek aangebied. Dit word bereik deur die uitbreiding van die klassieke Gabriel-hoofkomponent-analise bi stipping as 'n beginpunt te gebruik. 'n Simulasiestudie word aangebied om die sukses van die nuutontwikkelde metodologie beter te verstaan. Masters 2025-06-09T12:47:27Z 2025-06-09T12:47:27Z 2025-03 Thesis https://scholar.sun.ac.za/handle/10019.1/132480 en_ZA Stellenbosch University xvii, 91 pages : illustrations, includes annexures application/pdf Stellenbosch : Stellenbosch University |
| spellingShingle | Multivariate analysis -- Graphic methods Correspondence analysis (Statistics) Biplots Graphical modeling (Statistics) Multiple comparisons (Statistics) UCTD Ramaisa, Mokgeseng GPAbin to unify principal component analysis biplots from multiple imputations |
| title | GPAbin to unify principal component analysis biplots from multiple imputations |
| title_full | GPAbin to unify principal component analysis biplots from multiple imputations |
| title_fullStr | GPAbin to unify principal component analysis biplots from multiple imputations |
| title_full_unstemmed | GPAbin to unify principal component analysis biplots from multiple imputations |
| title_short | GPAbin to unify principal component analysis biplots from multiple imputations |
| title_sort | gpabin to unify principal component analysis biplots from multiple imputations |
| topic | Multivariate analysis -- Graphic methods Correspondence analysis (Statistics) Biplots Graphical modeling (Statistics) Multiple comparisons (Statistics) UCTD |
| url | https://scholar.sun.ac.za/handle/10019.1/132480 |
| work_keys_str_mv | AT ramaisamokgeseng gpabintounifyprincipalcomponentanalysisbiplotsfrommultipleimputations |