An Algorithm for the Multidimensional Analysis of the Overestimate and Underestimate of Property Rental Value

Multidimensional attributes related to a real estate property interfere with the cognitive ability of potential real estate property user to process all available information in order to determine the real rental market rate and make a fully informed decision. Although there are various web platforms for property leasing, it is difficult for a potential user to find two equivalent real estate properties and thus make relevant comparisons. This paper proposes a potential solution for the rental price estimate of a property to be leased, as well as scaling and comparison of its overestimated or underestimated value to all other available properties. This example is developed on the basis of real estate market in Sarajevo. Random Forest algorithm was used for the development of regression model. The principle aim of this paper is to offer an applicative module, based on Data Science methodology, for automatic scaling and the comparison of property overestimated or underestimated rental rates to other residential buildings and taking in consideration multidimensionality of their physical and geolocation attributes.