Interactive Student Performance Dashboard with Predictive Analytics

Design a Flask-based web application that aggregates student performance data using pandas for data processing. Implement machine learning predictions using scikit-learn to forecast potential student dropout risks based on historical academic performance, attendance, and engagement metrics. Include interactive visualizations with Plotly showing trend analysis, and create a feature that generates personalized intervention recommendations for at-risk students. Ensure the dashboard supports multiple data input formats and provides role-based access for administrators and educators.