surveydzofficial-oss/Algeria-laptop-price-prediction-website

# Algerian Laptop Price Predictor A Django web application that predicts laptop prices in the Algerian market using machine learning. ## Overview This project uses a trained Linear Regression model to predict laptop prices based on specifications like brand, series, CPU, GPU, RAM, storage, and condition. The model achieves 88.1% accuracy (R² score) with a Mean Absolute Error of 19,316 DZD. ## Features - Interactive web form for laptop specifications input - Dynamic dropdowns (brand→series, CPU brand→family) - Real-time price prediction with confidence intervals - Modern, responsive design with Algerian theme - Model performance: R² = 0.881, MAPE = 14.82%, RMSE = 34,390 DZD ## Usage 1. Navigate to the homepage 2. Fill in laptop specifications: - Basic info: Brand, Series, Condition, Year - CPU: Brand, Family, Generation, Suffix, Professional flag - GPU: Model, Suffix - Memory: RAM size, SSD size, HDD size 3. Click "Predict Price" to get estimation 4. View predicted price with min/max range based on model uncertainty ## Model Information The machine learning model was trained on Algerian laptop market data. While a Gradient Boosting model achieved higher performance during testing, we chose Linear Regression for production deployment due to its computational efficiency and suitability for CPU-only backend environments. **Linear Regression Model Performance:** - **R² Score**: 0.8810 (88.1% variance explained) - **MAPE**: 14.82% - **MAE**: 19,316 DZD - **RMSE**: 34,390 DZD The Linear Regression model provides an excellent balance between prediction accuracy and computational simplicity, making it ideal for real-time web applications without requiring GPU acceleration. Price ranges are calculated using ±RMSE to provide confidence intervals. ## Project Structure ``` predict_price/ ├── manage.py ├── predict_price/ # Django project settings ├── predictor/ # Main application │ ├── forms.py # Form definitions │ ├── views.py …