Automated Valuation Model based on fuzzy and rough set theory for real estate market with insufficient source data

Abstract Objective monitoring of the real estate value is a requirement to maintain balance, increase security and minimize the risk of a crisis in the financial and economic sector of every country. The valuation of real estate is usually considered from two points of view, i.e. individual valuation and mass appraisal. It is commonly believed that Automated Valuation Models (AVM) should be devoted to mass appraisal, which requires a large size of databases (wider knowledge) and automated procedures. These models, however, have a wider spectrum of application. The main aim of the study is to elaborate on a decision-making algorithm in the form of an Automated Valuation Model that uses the assumptions of the decision-making theory and data mining technology (Rough Set Theory (RST) and Value Tolerance Relation (VTR) - Fuzzy logic). The algorithm gives the opportunity to obtain the value of real estate where, using “if...then...” rules, we can account for the possibility of a non-deterministic relationship between real estate variables. It is applied to a small dataset of commercial real estate properties in Italy and residential ones in Poland. The proposed solution is universal and may be used in any other domain with imprecise and vague data.