Avaliação dos preços de imóveis na cidade de fortaleza, com a utilização de redes neurais artificiais, para a composição do ITBI

Abstract

There are several agents who study the full breadth of the real estate market. The value of housing and the variables that cause fluctuations in prices have been the subject of investigation by many specialists in different areas, such as economists, engineers, real estate investors, among others. In addition, there are also other players involved in a real estate transaction who are also interested in a more technical determination for the value of a property, such as real estate, banks, developers, builders, investors and insurance companies. Thus, the present study proposes to make a price evaluation in the city of Fortaleza, with the support of techniques of Artificial Neural Networks, for the composition of ITBI (Real Estate Transmission Tax). The data analyzed were provided by the Planning Department of the Municipality of Fortaleza and are related to the transactions of real estate transactions of vertical residential buildings, built in the period from 2012 to 2016. The evaluation model was developed from the selection of the main variables that influence the distribution of these prices. 14 input variables were used, which deal with the physical characteristics of the properties analyzed, including the location variable. The results of the survey were organized in the neighborhoods of Fortaleza with the highest number of real estate transactions registered in the last five years. Thus, it was possible to present the efficacy of this prediction model, with the largest contribution being support to the decision in estimating real estate prices for tax purposes.