Machine Learning Approaches for Analyzing and Predicting Concurrent Delays in Construction Projects

Construction projects are highly complex and often vulnerable to delays due to multiple overlapping factors. This study employs machine learning techniques to analyse and predict concurrent delays, a critical issue in the construction industry. A synthetic dataset was generated to represent key factors influencing delays, including labour hours, resource availability, weather conditions, and project complexity. Preprocessing steps such as normalization and outlier handling were applied to ensure data quality. K-means clustering was used to identify patterns in delay occurrences, while a Random Forest classifier provided predictive insights. The model demonstrated robust performance, achieving high accuracy, precision, and recall in delay prediction. Comparative analysis with traditional methods highlighted the advantages of machine learning in identifying complex relationships between variables. This study provides actionable insights for stakeholders to mitigate delays and optimize project outcomes using data-driven approaches.