A Novel LSTM-Based Approach to PERT Analysis in Construction Project Management

In order to anticipate Optimistic (O), Pessimistic (P), and Most Likely (M) values, this research study introduces a novel technique to PERT (Project Evaluation and Review Technique) analysis in construction projects using Long Short-Term Memory (LSTM) models. Recurrent neural networks (RNNs) of the Long Short-Term Memory (LSTM) variety are highly proficient in processing sequential input, which makes them perfect for modeling the intricate temporal connections present in building projects. For construction projects to be completed successfully, precise cost and schedule estimation is essential. The study focuses on project cost and schedule forecasting using LSTM, which is a crucial part of PERT analysis. Because the LSTM model can capture the long-term dependencies in project data, it can yield predictions that are more reliable and accurate than those produced by traditional statistical techniques. This research improves the field by demonstrating how LSTM may enhance project planning and management and offer a more sophisticated tool for decision-making in construction projects. This study provides a potential direction for further investigation and implementation by highlighting the significance of incorporating cutting-edge machine learning algorithms into sustainable construction project management.