🏠 Helsinki House Price Analysis

A comprehensive analysis and prediction system for property prices in the Helsinki metropolitan area, featuring data collection, exploratory analysis, and an interactive multi-page Streamlit application.

🌐 Live Demo

I've built an interactive demo to explore the results: 🔗 Live Application

📚 Research & Analysis

This project includes comprehensive Jupyter notebooks covering the entire data science pipeline:

• 📊 Web Scraping - Data collection from Etuovi.com
• 🧹 Data Cleaning - Raw data preprocessing and validation
• 🔍 Exploratory Analysis - Statistical analysis and geographical visualization

🚀 Interactive Application Features

📊 Data Explorer

• Interactive property map with multiple color schemes (percentile-based, logarithmic, price tiers)
• 3D density visualization showing property distribution
• Geographic analysis across Helsinki, Vantaa, and Espoo
• Real-time property statistics and insights

🔮 Price Predictor

• Machine learning-powered price predictions using Random Forest
• 25+ location presets covering the Helsinki metropolitan area
• Market comparison and contextual analysis
• Property characteristic impact visualization

🧠 Model Analysis

• Comprehensive performance metrics (R², MAE, RMSE, MAPE)
• Interactive feature importance analysis with error bars
• Partial dependence plots showing individual feature effects
• Model reliability assessment and insights

📁 Project Structure

helsinki-house-price/
├── app/
│   ├── 🏠_Home.py                      # Main landing page
│   └── pages/
│       ├── 1_📊_Data_Explorer.py       # Data visualization and exploration
│       ├── 2_🔮_Price_Predictor.py     # Price prediction interface
│       └── 3_🧠_Model_Analysis.py      # Model performance and analysis
├── notebooks/                          # Jupyter analysis notebooks
├── data/
│   └── cleaned/
│       └── helsinki_house_price_cleaned.xls
├── requirements.txt
└── README.md

🛠️ Local Installation

1. Clone the repository

   git clone https://github.com/albertonietos/helsinki-house-price.git
   cd helsinki-house-price

2. Create virtual environment

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies

   pip install -r requirements.txt

🚀 Running the Application

streamlit run app/🏠_Home.py

The application will open in your browser at http://localhost:8501

📊 Data Source

The dataset contains property listings scraped from Etuovi.com covering:

• Helsinki - City center and surrounding areas
• Vantaa - Airport region and suburbs
• Espoo - Tech hub and residential areas

Features Used

• Size: Property area in square meters
• Year: Year the property was built
• Total_rooms: Number of rooms
• Latitude/Longitude: Geographic coordinates
• Price: Target variable (in euros)

🤖 Machine Learning Model

• Algorithm: Random Forest Regressor
• Features: Size, Year, Total number of rooms, Latitude, Longitude
• Performance: ~80% R² score
• Validation: Train/test split

🎨 Visualization Features

Color Schemes

• Percentile-based: Balanced color distribution
• Logarithmic: Good for wide price ranges
• Price tiers: Quartile-based categories
• Linear: Traditional min-max scaling

Interactive Elements

• Hover tooltips with property details
• Zoom and pan capabilities
• Real-time parameter updates
• Responsive design

📈 Model Insights

The analysis reveals key factors affecting Helsinki property prices:

1. Size - Most important factor (~60% importance)
2. Location - Latitude/Longitude combined (~25% importance)
3. Year Built - Age and condition factor (~10% importance)
4. Room Count - Layout and functionality (~5% importance)

⚠️ Limitations

• Model based on limited feature set
• Market conditions change over time
• External factors not captured (schools, transport, etc.)
• Remember: All models are wrong, some models are useful

🔧 Technical Details

Dependencies

• Streamlit: Web application framework
• Pandas: Data manipulation and analysis
• NumPy: Numerical computing
• Scikit-learn: Machine learning algorithms
• Plotly: Interactive visualizations
• PyDeck: Advanced mapping capabilities

Performance Optimizations

• @st.cache_data for data loading
• @st.cache_resource for model training
• Efficient data processing pipelines
• Responsive UI design

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

📄 License

This project is open source and available under the MIT License.

👨‍💻 Author

Alberto Nieto

• LinkedIn: albertonietosandino
• GitHub: albertonietos

🙏 Acknowledgments

• Data sourced from Etuovi.com
• Built with Streamlit
• Visualizations powered by Plotly and PyDeck

---

For questions, suggestions, or issues, please open a GitHub issue or contact me.