No abstract

Purpose The paper aims to investigate the application of particle swarm optimisation and back propagation in weights optimisation and training of artificial neural networks within the mass appraisal industry and to compare the performance with standalone back propagation, genetic algorithm with back propagation and regression models. Design/methodology/approach The study utilised linear regression modelling before the semi-log and log-log models with a sample of 3,242 single-family dwellings. This was followed by the hybrid systems in the selection of optimal attribute weights and training of the artificial neural networks. Also, the standalone back propagation algorithm was used for the network training, and finally, the performance of each model was evaluated using accuracy test statistics. Findings The study found that combining particle swarm optimisation with back propagation in global and local search for attribute weights enhances the predictive accuracy of artificial neural networks. This also enhances transparency of the process, because it shows relative importance of attributes. Research limitations/implications A robust assessment of the models’ predictive accuracy was inhibited by fewer accuracy test statistics found in the software. The research demonstrates the efficacy of combining two models in the assessment of property values. Originality/value This work demonstrated the practicability of combining particle swarm optimisation with back propagation algorithms in finding optimal weights and training of the artificial neural networks within the mass appraisal environment.

The study examines the influence of four spatial weighting functions and bandwidths on the performance of geographically weighted regression (GWR), including fixed Gaussian and bisquare adaptive kernel functions, and adaptive Gaussian and bi-square kernel functions relative to the global hedonic ordinary least squares (OLS) models. A demonstration of the techniques using data on 3.232 house sales in Cape Town suggests that the Gaussian-shaped adaptive kernel bandwidth provides a better fit, spatial patterns and predictive accuracy than the other schemes used in GWR. Thus, we conclude that the Gaussian shape with both fixed and adaptive kernel functions provides a suitable framework for house price valuation in Cape Town.

The purpose of this study is to explore the need to adopt the CAMA (computer assisted mass appraisal) based system into the Nigerian property market with an example data from Nasarawa in Nasarawa State. Using the hedonic regression models on 312 samples, the results reveal the linear form to outperform the semi-log specification. This demonstration shows that CAMA system can be introduced into the local market of Nasarawa and indeed the whole of Nigeria for property taxes.

PurposeThe paper introduced the use of a hybrid system of neural networks support vector machines (NNSVMs) consisting of artificial neural networks (ANNs) and support vector machines (SVMs) to price single-family properties.Design/methodology/approachThe mechanism of the hybrid system is such that its output is given by the SVMs which utilise the results of the ANNs as their input. The results are compared to other property pricing modelling techniques including the standalone ANNs, SVMs, geographically weighted regression (GWR), spatial error model (SEM), spatial lag model (SLM) and the ordinary least squares (OLS). The techniques were applied to a dataset of 3,225 properties sold during the period, January 2012 to May 2014 in Cape Town, South Africa.FindingsThe results demonstrate that the hybrid system performed better than ANNs, SVMs and the OLS. However, in comparison to the spatial models (GWR, SEM and SLM) the hybrid system performed abysmally under with SEM favoured as the best pricing technique.Originality/valueThe findings extend the debate in the body of knowledge that the results of the OLS can significantly be improved through the use of spatial models that correct bias estimates and vary prices across the different property locations. Additionally, utilising the result of the hybrid system is thus affected by the black-box nature of the ANNs and SVMs limiting its use to purposes of checks on estimates predicted by the regression-based models.