STIMA DEGLI EFFETTI DELLA RIGENERAZIONE URBANA SUI VALORI IMMOBILIARI

Abstract

Urban regeneration interventions, increasingly implemented in response to the uncontrolled urbanisation of cities, can generate social, environmental, and economic benefits.  This study aims to investigate how urban regeneration influences the price of residential real estate. This paper compares techniques commonly used in practice, such as Multiple Linear Regression (MLR), and innovative Artificial Intelligence (AI) models like Artificial Neural Networks (ANNs). The analysis shows that some of the criticalities of MLR, such as the inability to handle non-linearity and collinearity between variables, can be overcome by resorting to AI algorithms. However, the latter fail – for instance – to evaluate the marginal prices of input variables. Therefore, the research first aims to provide a panel of variables useful for predicting real estate values following changes in the quality of the urban environment. Then, a methodology that involves the joint use of MLR and ANN is defined. This is to demonstrate that AI models, when supported by traditional models, can return a broader set of information to valuers and represent a more valid support to decision-making.