How helloharun/car-price-prediction-using-linear-regression works
Main system components
• Backend Web Service:
• Contains the machine learning model (best_model.joblib) along with its scaler (scaler.joblib) and additional dataset details (car_details.csv).
• Runs the Flask-based API (as indicated in ml_server.ipynb) which exposes the following endpoints:
▪ GET /sample_request (sample_request.py)
▪ GET /parameters_explanation (parameters_explanation.py)
▪ POST /home (home.py)
• Also may include auxiliary scripts and exploration notebooks (like main_models.ipynb) for model development and testing.
• Frontend Client:
• Implemented via Streamlit (client_app.py located in st_clients), which allows users to interact with the API by sending data and receiving predictions.
• Additional scripts in st_clients (like about_me.py) support supplementary information pages or client-side logic.
Important architectural patterns and technologies
• Clearly label that the backend is powered by Python, Flask and Scikit-learn. Indicate data manipulation handled by Pandas.
• Denote that Streamlit is used for the frontend rendering.
• Mention the use of joblib for model persistence.
• RESTful API design, which follows conventional endpoint structures.
Final advice
• Every component is clearly labeled to reflect its role. For instance, note which endpoints perform GET vs. POST operations, and ensure that the data flows (request/response cycle) are easy to follow.
• How the ML model is integrated into the RESTful API, which is then consumed by the client app.
• Don’t hesitate to include annotations explaining that the notebooks represent development artifacts rather than runtime components, but they are crucial for understanding the model’s training process.