How narukavya/student-performance-regression works
This repository predicts students’ final grades from prior grades and study-related features. Its workflow is: a user enters inputs in the Gradio app, which loads the saved regression model and returns a prediction. Separate scripts inspect the dataset, explore model predictions visually, and train/evaluate and save the model. The README is empty; relationships shown here are supported by the sampled scripts, not by documentation.
Main entry points: app.py Gradio prediction interface · student-mat.csv Student dataset · eda.py Prediction visualization · linear_regression.py Regression training and evaluation
Components 8
Data exploration
- Student dataset
data/student-mat.csv - Dataset checks
train.py - Prediction visualization
eda.py - Actual vs predicted plot
images
Model training
- Regression training and evaluation
linear_regression.py - Saved regression model
model.pkl
Prediction app
- Gradio prediction interface
app.py
Other components
- User
Connections 8
- Dataset checks → Student dataset: reads
train.py - Prediction visualization → Student dataset: reads
eda.py - Regression training and evaluation → Student dataset: reads
linear_regression.py - Prediction visualization → Actual vs predicted plot: saves plot
eda.py - Regression training and evaluation → Saved regression model: saves model
linear_regression.py - User → Gradio prediction interface: enters features
app.py - Gradio prediction interface → Saved regression model: loads and predicts
app.py - Gradio prediction interface → User: displays grade
app.py