Analysis of the capabilities of information and simulation mathematical modeling technologies for forecasting the cost of work at the construction stage

The work is devoted to the analysis and selection of methods of information and simulation mathematical modeling to improve the accuracy of forecasts of construction costs. It is determined that an essential factor in the successful application of methods of information and simulation mathematical modeling is the use of neural network technologies and machine learning capable of taking into account the influence of macroeconomic factors on the cost of construction in the region. It is shown that the effectiveness of models depends to a greater extent on the specifics of the data and their successful application in one region does not guarantee the achievement of similar forecasting results when assessing construction in another region. The analysis showed that it is most effective to use hybrid methods and models based on neural network technologies to forecast the cost of construction.