Utilizing Predictive Analytics and Machine Learning for Enhanced Project Risk Management and Resource Optimization

Effective risk management is critical for minimizing unforeseen costs and ensuring project success through proactive strategies. This study introduces an innovative real-time risk management framework leveraging predictive analytics and machine learning (ML). By analyzing historical project data, this approach identifies potential risks, emphasizing parameters such as task durations, resource allocation, and project outcomes. A t-distributed Stochastic Neighbor Embedding (t-SNE) technique optimizes feature selection, reducing dimensionality while retaining essential data properties. Model evaluation metrics include accuracy, precision, recall, and F1-score. The results indicate that the Gradient Boosting Machine (GBM) outperforms previous models, achieving 85% accuracy, 82% precision, 85% recall, and an 80% F1-score. Furthermore, predictive analytics significantly improves resource utilization efficiency (85%) and reduces project costs by 10%, compared to 70% and 5%, respectively, achieved by traditional methods. While GBM demonstrates superior overall performance, Logistic Regression (LR) offers favorable precision-recall trade-offs, underscoring the importance of tailored model selection in project risk management