In the field of civil engineering, there are some traditional methods of property evaluation that deal with these techniques. However, there is controversy about what would bring the best performance, a greater degree of ease and clarity without presenting multicollinearity. This controversy is due to the difficulty of finding appropriate predictive variables in real estate valuation since they often do not fit the binary model, involving human subjectivity. From this, the research aims to propose improvements in the property evaluation process with the use of artificial intelligence without presenting the effects of multicollinearity and autocorrelation, to predict the
value of the real estate market. The object of study is a standard 2-bedroom residential apartment with 48m-2 located in the central area of Jelgava, Latvia, in October 2019. Therefore, the methodology uses statistical inference as an initial analysis parameter and the fuzzy logic incorporates the best association rules which are originated from artificial intelligence extracted from the apriori algorithm. Finally, the results obtained by
regression and fuzzy were compared with the value in euros m-2, according to the official publication of the government of Latvia, referring to the market value of a 2-bedroom residential apartment in the city of Jelgava, Latvia, in October 2019, this government publication is the reference for this study. The statistical hypotheses
that allowed its validation were accepted. In the Fuzzy model, the results indicated an excellent equivalence to market prices in relation to the traditional valuation process.
Research on Factors Affecting Real Estate Values by Data Mining