Research on the Prediction of Construction Cost Based on BPNN

It is necessary to accurately predict the construction cost, especially in the early stage of the project. This paper adopts BPNN as a new artificial neural network method, which can effectively solve the problems existing in the current construction project cost prediction. BPNN method can capture the main influencing factors of construction cost and quickly estimate them. The cost of a project is affected by many factors. If these factors are treated as variables, it will not only increase the calculation workload, but also affect the accuracy of estimation. The main factor affecting the project cost is what affects the project cost. If the main cause is ignored, the accuracy of the forecast will be affected. This paper takes the cost prediction of residential projects as the research object. On this basis, it selects the main factors that affect the cost of real estate projects, analyzes and quantifies them, inputs them into the established BPNN model, takes the single cost of the project as its output set, and uses BPNN to predict the cost.