Empowering Young Adults: The Impact of Soft Skills Ventured through XGBoost and SHAP Analysis

The study, “Empowering Young Adults,” investigates the role of soft skills in ensuring students' confidence to excel in personal and professional life using advanced machine learning models. Initially, a total of 10 Machine Learning models and 3 Deep Learning models were trained out of which Gradient Boosting showed a high training accuracy of 93.3% but a low-test accuracy of 52.2%, leading to a huge gap between the two. Naive Bayes has the highest test accuracy of 69.9% with a training accuracy of 74.2%. To address this disparity as well as to improve the accuracies, SMOTE was applied, and XGBoost emerged as the best performer with a training accuracy of 91.1% and a test accuracy of 81.2%. The application of SHAP (SHapley Additive exPlanations) for visualization of the best model procured further explains and enumerates upon the importance of different soft skills This therefore highlights their significant impact on student confidence as well as the value of soft skills in educational outcomes, which demonstrates the efficacy of machine learning models in enhancing the assessment and development of these critical capacities.