Machine learning models for mass appraisals: advancing valuations in the digital era

Simple item page
dc.contributor.advisorDu Preez, Johan en_ZA
dc.contributor.authorDe Wet, Dominiqueen_ZA
dc.contributor.otherStellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.en_ZA
dc.date.accessioned2023-11-29T05:46:50Zen_ZA
dc.date.accessioned2024-01-08T17:25:08Zen_ZA
dc.date.available2023-11-29T05:46:50Zen_ZA
dc.date.available2024-01-08T17:25:08Zen_ZA
dc.date.issued2023-12en_ZA
dc.descriptionThesis (MEng)--Stellenbosch University, 2023.en_ZA
dc.description.abstractENGLISH ABSTRACT: A property appraisal is a professional, unbiased valuation that determines the market value of a property. Traditionally, property appraisals were conducted exclusively by professional appraisers. However, with increased data availability and enhanced computational resources, automated valuation models (AVMs) have gained widespread recognition as efficient tools to assist property appraisers in conducting mass appraisals. This report investigates the suitability of various machine learning (ML) methods as AVMs. The techniques include multiple polynomial regression, random forest regression, support vector regression, and a neural network. In addition to these four models, this report also introduces and assesses the advantages of a new innovative fusion model as an AVM. The fusion model is an ensemble approach that employs a neural network to combine the predictions from the four previous ML models, aiming to achieve improved accuracy and precision. This report uses property sales data from three specific neighbourhoods located within the Western Cape province of South Africa: Edgemead, Pinehurst, and Brackenfell. The results from this study indicate that all the individual ML techniques produce highly accurate property price predictions. However, they also yielded relatively large errors for some predictions. In contrast, the fusion model achieved greater accuracies and minimised most of its errors compared to the standalone models, establishing it as an effective AVM technique. This report provides a comprehensive framework for improving mass appraisal models.en_ZA
dc.description.abstractAFRIKAANSE OPSOMMING: ’n Eiendomswaardasie is ’n professionele, onpartydige skatting wat die markwaarde van ’n eiendom bepaal. Tradisioneel is eiendomswaardasies gewoonlik deur professionele waardeerders uitgevoer. Met verhoogde databeskikbaarheid en verbeterde berekeningshulpbronne het outomatiese waardasiemodelle (OWM) egter wydverspreide erkenning ontvang as doeltreffende instrumente om eiendomswaardeerders by te staan om massawaardasies uit te voer. Hierdie verslag ondersoek die geskiktheid van verskeie masjienleer (ML) tegnieke as OWMs. Die tegnieke sluit in meervoudige polinomiese regressie, willekeurige bos regressie, ondersteuningsvektor regressie en ’n neurale netwerk. Benewens hierdie vier modelle, word daar in hierdie verslag ook die voordele van ’n nuwe innoverende OWM samesmeltingsmodel bekend gestel. Die samesmeltingsmodel is ’n ensemblebenadering wat ’n neurale netwerk gebruik om die skattings van die vier vorige ML modelle te kombineer, met die doel om verbeterde akkuraatheid en presisie te bereik. Hierdie verslag gebruik eiendomsverkoopdata van drie spesifieke woonbuurte wat in die Wes-Kaap provinsie van Suid-Afrika gele¨e is: Edgemead, Pinehurst en Brackenfell. Die resultate van hierdie studie dui daarop dat al die individuele ML-tegnieke hoogs akkurate eiendomswaardasies lewer. Hulle het egter ook relatiewe groot foute vir sommige skattings opgelewer. Daarteenoor het die samesmeltingsmodel groter akkuraatheid behaal en die meeste van sy foute geminimaliseer in vergelyking met die selfstandige modelle, wat dit as die voorkeur-OWM-tegniek gevestig het. Hierdie verslag verskaf ’n omvattende raamwerk vir die verbetering van massa eiendomswaardasiemodelle.af_ZA
dc.description.versionMastersen_ZA
dc.format.extentix, 103 pages : illustrationsen_ZA
dc.identifier.urihttps://scholar.sun.ac.za/handle/10019.1/128963en_ZA
dc.language.isoen_ZAen_ZA
dc.language.isoen_ZAen_ZA
dc.publisherStellenbosch : Stellenbosch Universityen_ZA
dc.rights.holderStellenbosch Universityen_ZA
dc.subject.lcshMachine learning -- Western Cape (South Africa)en_ZA
dc.subject.lcshReal estate business -- Technological innovations -- Western Cape (South Africa)en_ZA
dc.subject.lcshProperty -- Valuation -- Western Cape (South Africa)en_ZA
dc.subject.lcshDigital computer simulation -- Western Cape (South Africa)en_ZA
dc.titleMachine learning models for mass appraisals: advancing valuations in the digital eraen_ZA
dc.typeThesisen_ZA
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
dewet_machine_2023.pdf
Size:
5.73 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Masters Degrees (Electrical and Electronic Engineering)