Prediction Accuracy for Property Tax Mass Appraisal: A Comparison Between Regularized Machine Learning and the Eigenvector Spatial Filter Approach

Prediction accuracy for mass appraisal has evolved substantially over the last few decades, facilitated by the revolution in data availability and the advancement of computational software. Accompanying these advances, newer geospatial approaches and machine learning algorithms have opened up new ho...

Full description

Saved in:
Bibliographic Details
Published inJournal of property tax assessment & administration Vol. 19; no. 2; pp. 21 - 46
Main Authors Mccord, M.J, Davis, P.T, Bidanset, P.E, Hermans, L.D
Format Journal Article
LanguageEnglish
Published Kansas City International Association of Assessing Officers 01.07.2022
Subjects
Accuracy
Ad valorem taxes
Analysis
Artificial intelligence
Automation
Bias
Data mining
Eigenvectors
Generalized linear models
Geographic information systems
International taxation
Machine learning
Methods
Neural networks
Property taxes
Property values
Real property
Regression analysis
Software
Tax assessments
Taxation
Valuation
Indiana
United Kingdom
United Kingdom > UK
United States > US
Online AccessGet full text

Cover

More Information
Summary:Prediction accuracy for mass appraisal has evolved substantially over the last few decades, facilitated by the revolution in data availability and the advancement of computational software. Accompanying these advances, newer geospatial approaches and machine learning algorithms have opened up new horizons for price prediction and mass appraisal assessment. The application of machine learning (ML) and artificial intelligence (AI) within mass appraisal has generated considerable debate; these methods are often perceived as impractical because their explainability and defensibility--required for assessment application, notably in challenge scenarios--are limited. This study compares a traditional multiple regression analysis (MRA) approach with regularized (penalized) machine learning approaches and a more nuanced geostatistical technique, the eigenvector spatial filter (ESF) approach, applying data sets for two urban residential areas in the United Kingdom and the United States. The findings show the efficacy of the geostatistical ESF technique against the ML approaches--both of which outperform the traditional MRA. The findings also show the ESF approach provides the basis of a more understandable alternative spatial method for mass appraisal aligned with the MRA approach, with the spatial filters easily incorporated as predictors into MRA to alleviate spatial autocorrelation. Further, the penalized ML regression approaches offer a more practical alternative to other forms of ML for assessors. Both methods produce reliable yet understandable regression models for mass appraisal assessment.
ISSN:1357-1419
1548-3606