Classification Modelling for the Impact of Historical Site on Residential Property Rental Value in Osogbo, Nigeria

Abstract

Previous studies have revealed that historical sites have impact on property value. However, none of the previous studies have forecasted the impact of historical site on residential property value using a classification model. Data for the study were gathered from the record of recent letting in the study area. For the purpose of precision, this study adopted artificial neural network, logistic regression and support vector machine as model of classifying the rental value of residential property in Osogbo, Nigeria. The study considered relevant variables which include distance to cultural site, age of building, state of exterior and state of interior as input variables. Findings from the study revealed that the three adopted forecasting models have over 80% percent of the forecasted properties correctly classified which make the property forecasting reliable.

Rent index forecasting through neural networks

“…As compared to house price forecasting, research on rental price forecasting through machine learning (e.g. Clark and Lomax, 2018; Embaye et al , 2021; Hu et al , 2019; Li, 2018; Li and Li, 1996; Ma et al , 2018; Ma and Liu, 2019; Ming et al , 2020; Odubiyi et al , 2019; Oshodi et al , 2020, 2021; Oyedeji Joseph et al , 2018; Oyedeji and Oyewale, 2018; Rafatirad, 2017; Tsai and Pan, 2014; Wang and Cao, 2019; Zhang et al , 2019) seems relatively scare. Hu et al.…”

“…As compared to house price forecasting, research on rental price forecasting through machine learning (e.g. Clark and Lomax, 2018;Embaye et al, 2021;Hu et al, 2019;Li, 2018;Li and Li, 1996;Ma et al, 2018;Ma and Liu, 2019;Ming et al, 2020;Odubiyi et al, 2019;Oshodi et al, 2020Oshodi et al, , 2021Oyedeji Joseph et al, 2018;Oyedeji and Oyewale, 2018;Rafatirad, 2017;Tsai and Pan, 2014;Wang and Cao, 2019;Zhang et al, 2019) seems relatively scare. Hu et al (2019) explore the random forest, extra-trees, gradient-boosting, support vector, multi-layer perceptron neural network and k-nearest neighbor when building housing rent prediction models for Shenzhen in China in October 2017 and February 2018 and find that all of these algorithms, except for the support vector, generally present good performance with the random forest and extra-trees being the leaders.…”

Exploring Spatial Pattern of Residential Property Value around Cultural Heritage Sites in Ile-Ife, Nigeria

“…In addition to neighbourhood, locational and structural characteristics, plentiful commentators (e.g. Adegoke & Adebara, 2019;Armitage & Irons, 2005;Franco & Macdonald, 2016;Lazrak et al, 2014;Moro et al, 2013;Oyedeji, 2018) have established that the presence of cultural heritage and the distances from cultural heritage sites can influence the value of property. In summary, the factors determining property value can be categorised under three (3) broad factors; first the physical attributes of the property, the location of the property and the characteristics of the neighborhood where the property is located.…”

“…The study of Oyedeji (2018) developed a model for predicting rental values of residential properties proximate to historical site. The historical site examined was Osun-Osogbo groove, in Nigeria.…”

“…Moreover, the types of cultural heritages in developed economies may be different from those available in Nigeria and other developing economies. On the Nigerian scene, some studies (Adegoke & Adebara, 2019;Oyedeji, 2018) have investigated the impact of cultural heritages on property value. Yet, much is still unknown about the spatial pattern of the effects of cultural heritage sites on rental value of residential properties.…”