Forecasting the completion time of construction projects using the moving average method

The importance of implementing projects 'on time and budget' has increased due to the increasing number of projects in organisations. Various models have been introduced, such as earned value management (EVM), earned schedule management (ESM), earned duration management (EDM), and earned resource management (ERM), to monitor project performance throughout the project execution phase. These models offer various metrics to evaluate project performance regarding cost, time, and resource productivity. However, using different data science methods to estimate project duration has been an interesting topic. This study presents a new model using project data obtained during the project execution phase and regression models to ensure a more accurate estimate of activities and overall project duration. The mean absolute percentage error (MAPE) metric, commonly employed in most proposed models, was used to check the accuracy. Four projects were selected according to the information required in the proposed model. The results indicated that the mean prediction error for the selected project was 6.61%, which is lower than other proposed methods. The results also showed that the proposed model provided suitable predictions for both activities and project duration.