Comparative Study of Explainable AI Techniques for Student Placement Prediction

In recent years, it has increasingly been difficult to secure ideal employment without advanced knowledge and skills. Conventional methods of student placement analysis, based on manual handling of data along with basic statistical tools, are usually not capable for capturing complexities that influence employability outcomes. Hence, this paper creates a system to analyze student placement history through machine learning algorithms with Explainable Artificial Intelligence approaches. It first gathers and preprocesses a multi-class dataset of past placements to train several kinds of machine learning models, out of which the highest-performing model will be chosen according to accuracy metrics. The selected model is analyzed using SHAP and LIME to identify and contrast the main features influencing the placement outcomes. The comparative analysis leads to providing actionable insights in relation to feature importance, considering pros and cons of both methods. These results can empower educational institutions to identify critical factors of employability, optimize placement strategies, and improve students' career outcomes in the competitive labor market.