The Impact of Risk Assessment on Project Success based on Data-Driven Approach

Project success and sustainability are profoundly affected by risk assessment, which is why it is an essential component of project management. Background: This study takes a data-driven approach to studying how risk assessment affects project performance. It uses machine learning to improve risk categorization and mitigation. Methodology: Software project risk data is collected, preprocessed, and normalized. Feature selection refines the dataset for model efficiency. A Support Vector Machine (SVM) classifier, optimized with Bayesian techniques, is compared with Logistic Regression (LR) and Gradient Boosting Machine (GBM). Results: SVM outperforms LR and GBM, achieving 0.89 accuracy, 0.88 precision, 0.91 sensitivity, and 0.89 F1-score. From 409 risk variables, 207 factors are categorized into 27 key IT project risks. Data shows that 72.22% of project managers frequently identify risks, with 46.30% classifying them as "Always" significant. Conclusion: Findings emphasize hybrid risk management strategies for continuous risk monitoring and proactive mitigation, ensuring improved project success.