# π Dubizzle Egypt Car Price Prediction β ML Pipeline, App & Dashboard
A comprehensive data science and machine learning pipeline designed to predict used car prices from **Dubizzle Egypt** listings (available at
dubizzle.com.eg). This project encompasses **data scraping**, **exploratory data analysis (EDA)**, **preprocessing**, **model training**, a professional **Streamlit web app**, and an interactive **Power BI dashboard** for price prediction and market insights.
---
## π Project Structure
| File / Folder | Description |
|-------------------------------------------|-------------|
| `Dubizzle_Scraping.py` | Python script utilizing `requests` and `BeautifulSoup` to scrape car listing links and raw data from Dubizzle Egypt. |
| `dubizzle_full_dataset.csv` | Raw dataset post-scraping, potentially containing missing or inconsistent values. |
| `Dubizzle_Preprocessing.ipynb` | Jupyter notebook for data cleaning, null handling, outlier management, feature engineering (e.g., log transformations), and encoding. |
| `dubizzle_cleaned_dataset.csv` | Processed dataset optimized for modeling and app deployment. |
| `Dubizzle_Modeling.ipynb` | Notebook training multiple regression models (Linear, Ridge, Random Forest, etc.) with GridSearchCV, evaluated via RΒ², RMSE, and residuals. |
| `LinearRegression_model.pkl` | Serialized Linear Regression model saved using `pickle`. |
| `Deployment.py` | Streamlit app featuring three main pages: Home (KPIs), Visualizations (Plotly-based), and Price Prediction. |
| `dubizzle-cars-logo.png` | Logo integrated into the Streamlit app homepage. |
| `PowerBI_Dashboard/` | Contains Power BI dashboard file (`.pbix`) and visual exports for 3 report pages. |
---
## π Power BI Dashboard
An interactive dashboard built with **Power BI** for visualizing car market tr β¦