Project Title

A compact, reproducible ML pipeline for training and serving a supervised model (student-performance style dataset).

Overview

This project demonstrates an end-to-end machine learning workflow:

• Data ingestion and preprocessing
• Exploratory Data Analysis (notebooks)
• Model training, evaluation, and artifact storage
• Prediction pipeline and a minimal web UI for inference

Repository layout

• app.py — Minimal Flask app / entrypoint for serving predictions
• artifacts/ — saved datasets and model artifacts
• logs/ — runtime and training logs
• notebook/ — EDA and model training notebooks
• src/ — core source code
  • components/ — data_ingestion.py, data_transformation.py, model_trainer.py
  • utils.py, logger.py, exception.py — helpers and utilities
• templates/ — HTML templates for web UI

Quick start

1. Create and activate a Python environment (recommended Python 3.8+).
2. Install dependencies:

pip install -r requirements.txt

3. Run training pipeline:

python src/pipeline/train_pipeline.py

4. Run prediction server:

python app.py

5. Open the UI at http://localhost:5000 (if app.py starts Flask).

Usage

• Use the notebooks in notebook/ for EDA and model experiments.
• Put raw datasets in artifacts/ (or configure paths in src/components).
• Trained models and metrics are stored under artifacts/ for reproducibility.

Notes & Next steps

• Add detailed hyperparameters and evaluation metrics to the training outputs.
• Add unit tests for pipeline components.
• Containerize with Docker for consistent deployment.

If you'd like, I can also add example API requests, CI steps, or commit this change.

end to end machine learning