Machine Learning Student Dropout Risk Prediction Model

Develop a predictive JavaScript model using TensorFlow.js to calculate early warning indicators for student dropout risk in an online learning platform. Create a machine learning pipeline that ingests historical student interaction data, including login frequency, assignment completion rates, time spent in modules, and discussion forum participation. Build a neural network that generates risk scores and provides interpretable insights for academic advisors, with a React frontend for displaying probabilistic predictions and recommended interventions.