Machine Learning-Based Dynamic Cost Estimation Model for Construction Projects

Construction project management involves the implementation of the right methodologies in the right approach to complete the project within the planned deliverables. But, the actual practice of construction project management is challenging due to the dynamic work environment. Thereby in most situations, the planned expenditure is exceeded before the project completion stage. The necessity of vibrant cost estimation methodology is highly imperative while estimating the project cost while considering the possible uncertainties. The proposed research work aimed to develop an efficient cost estimation modal to determine the project cost with various conditions. The cost estimation model was developed as a machine learning optimization model using a Python programming application. The model was developed and evaluated using the actuals from two construction projects. The evaluation indicators like random forest ensure the validity of proposed model with almost 98% of accuracy in estimating the project cost. The proposed model could be used by just replacing the input variables for any category of a construction project.