A Comparison of Regression and Artificial Intelligence Methods in a Mass Appraisal Context

Abstract

This paper describes a comparative study where several regression and artificial intelligence (AI)-based methods are used to assess properties in Louisville, Kentucky. Four regressionbased methods [traditional multiple regression analysis (MRA), and three non-traditional regression-based methods, Support Vector Machines using sequential minimal optimization regression (SVM-SMO), additive regression, and M5P trees], and three AI-based methods [neural networks (NNs), radial basis function neural network (RBFNN), and memory-based reasoning (MBR)] are applied and compared under various simulation scenarios. The results indicate that non-traditional regressionbased methods perform better in all simulation scenarios, especially with homogeneous data sets. AI-based methods perform well with less homogeneous data sets under some simulation scenarios.