Determining the significant input parameters of a forecasting model for material resources of residential construction projects at the investment feasibility assessment stage

Abstract

Any investor/developer spends considerable time to make a construction decision in analyzing various parameters, such as the building site area, height restrictions, total area, quantity of necessary materials, etc. First of all, this is due to a high level of uncertainty and the complexity of estimating costs for the initial stages of project implementation. The unavailability of detailed design data at the investment justification stage makes it impossible to conduct a sufficiently accurate assessment of the value and period of construction to make highly accurate managerial decisions. However, one of the key issues, namely determining the significant parameters that have the greatest impact on future project-related decisions, has not been studied sufficiently. This article identifies the most significant parameters that impact the future characteristics of residential buildings based on an expert assessment using the a priori ranking method. The reliability of the assessment is confirmed using the coefficient of concordance. A conclusion is provided about their applicability as the input values of a forecasting model that uses machine learning and artificial intelligence to determine various technical and economic characteristics of residential buildings.