Research on Project Cost Estimation Model based on Machine Learning and Neural Network

Due to the lack of design depth of the general contract project, the cost estimate cannot be accurately estimated, and the cost estimate will directly affect each stage of the subsequent implementation of the project. Therefore, how to accurately predict the estimated cost becomes the key problem. In this paper, housing projects are selected as the research object, three kinds of machine learning models are compared, engineering features are taken as parameters, engineering feature vectors are extracted, and historical data are used for training. It is concluded that the estimation accuracy of the neural network model is the highest. Machine learning accurately predicts the cost estimate value, which helps to convert the general contract project into advance control, and realize the dynamic management of the project cost.