Prediktivni modeli u masovnoj procjeni vrijednosti nepokretnosti: izazovi i mogućnosti višestrukog linearnog regresionog modela
Predictive Models in Mass Real Estate Valuation: Challenges and Opportunities of the Multipl Linear Regression Model

Summary/Abstract: This paper explores the application of the multiple linear regression model in mass real estate valuation. The research was conducted on a sample of 527 apartments in the Republic of Srpska, using SPSS software version 23 for model implementation and result evaluation. In the context of dynamic market conditions and limited data availability, multiple linear regression enables a more accurate and objective market analysis compared to traditional methods. This technique allows for modeling complex relationships between factors affecting real estate prices, reducing subjectivity and errors inherent in traditional approaches. Although multiple linear regression is not a new method, it represents a significant advancement in the mass real estate valuation process, providing a solid foundation for work in developing markets. By comparing the estimated values with actual market prices and contract-defined prices, it was found that deviations are minimal, indicating the model’s high accuracy. Location, area, building age, and floor level were identified as key factors influencing real estate prices. The originality of this paper lies in the application of multiple linear regression under the specific market conditions of the Republic of Srpska, with a careful selection of variables reflecting local market specifics. Although artificial intelligence is increasingly becoming a key element in modern real estate valuation methods, this paper confirms that multiple regression analysis retains undeniable importance in real estate market analysis. Furthermore, the paper provides a solid foundation for further research and methodological improvement, with the potential for integrating advanced techniques such as machine learning algorithms. This opens the door to even more accurate predictions and more efficient application in market analyses.

• Details
• Contents