Civil engineering inspection for real estate evaluation with the use of artificial learning algorithms and fuzzy logic

Surgelas, V., Latvia Univ. of Life Sciences and Technologies, Jelgava . Faculty of Environment and Civil Engineering; Arhipova, I., Latvia Univ. of Life Sciences and Technologies, Jelgava . Faculty of Information Technologies; Pukite, V., Latvia Univ. of Life Sciences and Technologies, Jelgava . Faculty of Environment and Civil Engineering

Civil engineering inspection for real estate evaluation with the use of artificial learning algorithms and fuzzy logic

2020

Surgelas, V., Latvia Univ. of Life Sciences and Technologies, Jelgava (Latvia). Faculty of Environment and Civil Engineering | Arhipova, I., Latvia Univ. of Life Sciences and Technologies, Jelgava (Latvia). Faculty of Information Technologies | Pukite, V., Latvia Univ. of Life Sciences and Technologies, Jelgava (Latvia). Faculty of Environment and Civil Engineering

The technical inspection of a building carried out by an expert in civil engineering can identify and classify the physical conditions of the real estate; this generates relevant information for the protection and safety of users. Given the real conditions of the property, and for the real estate valuation universe, using artificial intelligence and fuzzy logic, it is possible to obtain the market price associated with the physical conditions of the building. The objective of this experiment is to develop a property evaluation model using a civil engineering inspection form associated with artificial intelligence, and fuzzy logic, and also compare with market value to verify the applicability of this inspection form. Therefore, the methodology used is based on technical inspection of civil engineering regarding the state of conservation of properties according to the model used in Portugal and adapted to the reality of Latvia. Artificial intelligence is applied after obtaining data from that report. From this, association rules are obtained, which are used in the diffuse logic to obtain the price of the apartment per square meter, and for comparison with the market value. For this purpose, 48 samples of residential apartments located in the city of Jelgava in Latvia are used, with an inspection carried out from October to December 2019. The main result is the 9% error metric, which demonstrates the possibility of applying the method proposed in this experiment. Thus, for each apartment sample consulted, it resulted in the state of conservation and a market value associated.

Afficher plus [+]