Mohammedkaif4666/Anomaly-Detector-Fee-Default-Predictor-AI-4-Kalnet- — reverse-engineered prompt

Build me a school admin web app called KALNET AI 4 that helps staff spot students who may need attention early.

I want a simple landing page, a student directory, and an admin dashboard. The dashboard should show attendance risk alerts, like students whose attendance suddenly dropped, and fee payment risk alerts, like students who may miss the next term payment. Use generated sample school data for around 500 students, including daily attendance, fee history, income bracket, transport use, siblings, and outstanding amount.

The app should train basic machine learning models with scikit learn only, no paid APIs. One model should find unusual attendance patterns and give a clear risk score. Another should predict whether a student is likely to pay on time, pay late, or default. Show the results in a clean interface that a school administrator could understand without technical knowledge.

Please include the backend API, dashboard pages, model training scripts, sample data generation, saved models, and a short evaluation report. Make it runnable locally and ready to deploy if needed.