In Russia, a neural network has learned to find long construction projects.
With the help of this system the housing development institute "Dom.RF" was able to determine the likely failures of housing commissioning dates, according to the press service of the organization. Earlier, Russian scientists created algorithms for a neural network, with the help of which it is possible to predict the harvest. "Dom.RF" uses machine learning algorithms based on the Unified Information System of Housing Construction (UISZhS). As the basis were taken standard algorithms of regression analysis, algorithms of artificial intelligence and neural networks, said the press service of the Institute. Grigory Gryaznov, the head of the Analytical Services of the UISZhS noted that the UISZhS database includes data on the ongoing housing construction projects, developers' declarations, progress of works and so on. On their basis the neuronet assesses each house and determines the probability of turning the construction into a long-lasting construction. In May, the Ministry of Construction stated the desire to use a neural network to assess the deterioration of buildings. The project uses artificial intelligence algorithms to analyze and communicate data on the need for repairs to the building. This spring, Russian scientists shared information about the development of an agricultural plant monitoring system based on neural network algorithms. The new tool should make it easier for farmers to predict crop yields and track seasonal problems.