A Fully Automated Adjustment of Ensemble Methods in Machine Learning for Modeling Complex Real Estate Systems

Abstract

The close relationship between collateral value and bank stability has led to a considerable need to a rapid and economical appraisal of real estate. The greater availability of information related to housing stock has prompted to the use of so-called big data and machine learning in the estimation of property prices. Although this methodology has already been applied to the real estate market to identify which variables influence dwelling prices, its use for estimating the price of properties is not so frequent. The application of this methodology has become more sophisticated over time, from applying simple methods to using the so-called ensemble methods and, while the estimation capacity has improved, it has only been applied to specific geographical areas. The main contribution of this article lies in developing an application for the entire Spanish market that fully automatically provides the best model for each municipality. Real estate property prices in 433 municipalities are estimated from a sample of 790,631 dwellings, using different ensemble methods based on decision trees such as bagging, boosting, and random forest. The results for estimating the price of dwellings show a good performance of the techniques developed, in terms of the error measures, with the best results being achieved using the techniques of bagging and random forest.

A Hybrid Approach for Mass Valuation of Residential Properties through Geographic Information Systems and Machine Learning Integration

“…Mass valuation can be defined as analyzing a set of factors to estimate the value of a large number of properties using statistical methods and standards (IAAO 2013a). There are a wide range of mass valuation methods like MRA (Zurada, Levitan, and Guan 2011; Benjamin, Guttery, and Sirmans 2020; Yilmazer and Kocaman 2020), Hedonic Pricing (Peterson and Flanagan 2009; Lisi 2019; Yamani, Ettarid, and Hajji 2019), Nominal Valuation (Yomralioglu 1993; Yomralioglu and Nisanci 2004; Mete and Yomralioglu 2019), Geographically Weighted Regression (GWR) (Huang, Wu, and Barry 2010; Dimopoulos and Moulas 2016; Wang, Li, and Yu 2020), Ensemble Methods (Alfaro‐Navarro et al 2020; Aydinoglu, Bovkir, and Colkesen 2021; Gnat 2021), and ANN (Demetriou 2017; Lee 2022; Yalpir 2018). With the improvements in computing power, GIS and Artificial Intelligence, Computer‐Assisted Mass Appraisal (CAMA) applications have become widespread and Automated Valuation Models (AVMs) have been adopted in many countries (Wang and Li 2019; Renigier‐Biłozor et al 2022).…”

House price prediction using hedonic pricing model and machine learning techniques

“…For applied machine learning, many algorithms exist. Nevertheless, eXtreme Gradient Boosting (XGB) is recently dominating the field due to its design for speed and model performance 46,47 . It can be used for both classification and prediction models alike.…”

“…Nevertheless, eXtreme Gradient Boosting (XGB) is recently dominating the field due to its design for speed and model performance. 46,47 It can be used for both classification and prediction models alike. It is an implementation of the gradient boosted decision trees.…”

Graph Convolutional Network-Based Model for Megacity Real Estate Valuation

“…There is another type of ensemble algorithm in which there is a strong dependency between individual weak learners that must be generated serially, which is represented by the Boosting algorithm. In their simulation of the Spanish real estate market, J.-L. Alfaro-Navarro et al [19]unearthed that the boosting algorithm outperformed the individual tree approach, though overall the random forest approach had moderately superior performance.In addition, J.-L. Alfaro-Navarro pointed out that ensemble learning methods tend to be applied in a limited way to specific geographic areas, while the best models tend to differ from city to city. The combination of decision trees and boosting ideas gave birth to the GBDT algorithm, which inherits the advantages and improves the disadvantages of decision trees and boosting, and, in turn, solves the problem of overfitting well by integrating multiple decision trees through the gradient boosting method.Meanwhile, the dilemma of sequential training and the difficulty of parallelism common to boosting algorithms has been effectively resolved with the advancement of XG-Boost and LightGBM (a framework for implementing the GDBT algorithm).…”