MazenSr/Egypt-Used-Car-Price-Predictor

AI-powered used car price prediction application built with XGBoost, Scikit-learn, and Streamlit, featuring intelligent price estimation and interactive data visualizations. # 🚗 Egypt Used Car Price Predictor A machine learning-powered web application that estimates the market value of used cars in Egypt based on vehicle specifications and provides interactive market insights. **🌐 Live Demo:** egypt-used-car-price-predic… --- ## Preview ### Prediction Tab ### Market Insights Tab --- ## Features - 🚗 Predict used car prices instantly - 📊 Interactive market analysis - 📈 Historical price trends by manufacturing year - 📉 Mileage vs. Price visualization - 📦 Price distribution analysis - 🎨 Automatic default values based on the selected Brand & Model - ⚡ Fast inference using a trained machine learning pipeline - ✨ Clean user-friendly interface --- ## Vehicle Specifications The prediction model uses: - Brand - Model - Manufacturing Year - Mileage (km) - Color - Automatic Transmission - Air Conditioner - Power Steering - Remote Control --- ## Machine Learning Pipeline The project follows a complete **End-to-End machine learning workflow**, covering every stage of a typical machine learning project, including: - Data Cleaning - Exploratory Data Analysis (EDA) - Custom Feature Engineering - Data Preprocessing - Hyperparameter Optimization - Learning Curve Analysis - Model Evaluation - Model Interpretation (Permutation Feature Importance) - Residual and Error Analysis - Model Serialization The final model is trained using **Scikit-learn pipelines**, allowing raw vehicle information to be transformed automatically into accurate price predictions. --- ## Market Insights Besides predicting prices, the application provides: - Dataset vehicle count - Minimum price - Maximum price - Current model prediction - Price trends across years - Mileage vs. Price relationship - Price distribution histogram --- ## Dataset - **Source:** Hatla2ee used car listings - **Dataset:** This project uses the Kaggle version of the dataset; the data scraping was performed by the dataset publisher …