How hfzhimn/student-dropout-prediction-app works
This project is an early-warning prototype for identifying students at risk of dropping out. Its interactive workflow collects five student attributes in a Streamlit app, passes them to a loaded Random Forest model, and displays a risk message. The repository also includes a data-science notebook and dataset; the README describes preprocessing, exploratory analysis, and model development, plus a separate Tableau monitoring dashboard. The notebook’s implementation and the dashboard itself were not inspected, so their internal behavior and connections beyond documented capabilities are not inferred.
Main entry points: app.py Prediction app · notebookDSAkhir.ipynb Data preprocessing · data.csv Student dataset · model.joblib Random Forest model
Components 10
Data Analysis
- Student dataset
data.csv - Data preprocessing
notebookDSAkhir.ipynb - Exploratory analysis
notebookDSAkhir.ipynb
Risk Prediction
- Model development
notebookDSAkhir.ipynb - Random Forest model
model.joblib - Prediction app
app.py - Risk prediction
app.py
Business Monitoring
- Tableau dashboard
Other components
- Institute staff
- Staff app user
Connections 7
- Student dataset → Data preprocessing: provides data
README.md - Data preprocessing → Exploratory analysis: prepares data
README.md - Exploratory analysis → Model development: informs modeling
README.md - Staff app user → Prediction app: enters attributes
app.py - Prediction app → Random Forest model: loads and predicts
app.py - Prediction app → Risk prediction: displays result
app.py - Institute staff → Tableau dashboard: monitors performance
README.md