machine learning
Car Price Prediction Web Application
Problem
Used-car records needed a repeatable modelling pipeline and an accessible interface for estimating vehicle prices.
Main result
Connected a validated machine-learning pipeline to a FastAPI backend and a browser-based prediction interface.
Project work
- Data preparation and feature engineering
- Categorical encoding and price transformation
- Train/test validation
- Comparison of regression models
- MAE and RMSE evaluation
- FastAPI prediction endpoint
- Jinja2 and HTML/CSS prediction interface
Technologies
PythonPandasNumPyscikit-learnFastAPIJinja2HTML/CSS