A spatially based artificial neural network mass valuation model for land consolidation

Abstract

Land consolidation, which aims to promote sustainable development of rural areas, involves the reorganization of space through land reallocation, both in terms of ownership and land parcel boundaries. Land reallocation, which is the core part of such schemes, is based on land values because each landowner is entitled to receive a property with approximately the same land value after land consolidation. Therefore, land value, which in the case of Cyprus is the market value, is a critical parameter, and hence it should be reliable, accurate, and fairly valued. However, the conventional land valuation process has some weaknesses. It is carried out manually and empirically by a five-member Land Valuation Committee, which visits every unique parcel in the consolidated area to assign a market value. As a result, it is time consuming and hence costly. Moreover, the outcomes can be inconsistent across valuators for whom, in the case of such a mass appraisal procedure, it is hard to analytically calculate the scores for a series of land valuation factors and compare all of these for hundreds of land parcels using a manual process. A solution to these shortcomings is the use of automated valuation models. In this context, this paper presents the development, implementation, and evaluation of an artificial neural network automated valuation model combined with a geographical information system applied in a land consolidation case study area in Cyprus. The model has been tested for quality assurance based on international standards. The evaluation showed that a sample of 15% of the selected land parcel values provided by the Land Valuation Committee is adequate for appraising the land values of all parcels in the land consolidation area with a high or acceptable accuracy, reliability, and consistency. Consequently, the automated valuation model is highly efficient compared to the conventional land valuation method since it may reduce time and resources used by up to 80%. Although the new process is based partly on the Land Valuation Committee sample, which inherently carries inconsistencies, it is systematic, analytical, and standardized, hence enhancing transparency. The comparison of artificial neural networks with similar linear and nonlinear models applied to the same case study area showed that it is capable of producing better results than the former and similar outcomes to the latter.

Mass Appraisal Models of Real Estate in the 21st Century: A Systematic Literature Review

“…The ANN performs well for modeling the non-linear relationship because of its characteristics of semi-parametric regression. In addition to the basic MRA, although researchers have to face the "black box" of the ANN's structure, it is still the most popular model used in AI-based models [36][37][38][39][40][41][42][43][44][45][46][47][48][49].…”

“…The spatial characteristics together with its non-spatial information contribute to the property value. Many scholars have paid attention to the GIS features of properties and considered their impacts on appraisal [44,[78][79][80][81][82][83][84][85][86]. Meanwhile, some scholars such as Bourassa et al (2007) [87], McCluskey and Borst (2007) [88] have made improvements in the classification and compilation of GIS-based model.…”

Land valuation in support of responsible land consolidation on Ghana’s rural customary lands

“…A further strength of the approach stemming from the assignment of weight in the process is the transparency and participation it affords the local people. Other approaches that use AVMs as suggested by Brankovicé t al., (2015) and Demetriou (2016a) that rather use multiple regression and geographically weighted regression, present approaches that do not involve the local people. They also present cases that cannot be analytically explained to local farmers.…”

Determining the effect of land consolidation on agricultural production using a novel assessment framework

“…As an important part of the theory and practice of LC, evaluating LC effectiveness has increasingly drawn great attention from academics and government (Chartin et al, 2013;Demetriou, 2017;Janus & Markuszewska, 2017;Lerman & Cimpoies, 2006;Zheng et al, 2017). Specifically, changes in the agricultural production caused by LC have been counted as an important indicator in evaluating the effectiveness of LC.…”