Pedestrian-Friendly Areas in Cologne: SVM Analysis

Overview

This project uses Support Vector Machines (SVM) to classify pedestrian-friendly areas in Cologne, Germany, based on real-world data fetched from OpenStreetMap (OSM). The analysis incorporates pedestrian, traffic, and cycleway routes and visualizes the decision boundary, feature importance, and routes on a map.

Table of Contents

1. Overview
2. Data Sources
3. Features
4. Installation
5. Usage
6. Visualizations
7. Results
8. Contributing
9. License

Data Sources

The data for this project is fetched directly from OpenStreetMap using the osmnx library. Key data includes:

• Pedestrian routes (highway=footway, highway=pedestrian, highway=path)
• Traffic roads (highway=motorway, highway=primary, highway=secondary, highway=tertiary)
• Cycleways (highway=cycleway)

Features

1. Fetch Data from OpenStreetMap: Retrieves all pedestrian routes, traffic roads, and cycleways for Cologne.
2. Simulated Neighborhood Data: Generates 100 data points with realistic variations for training.
3. SVM Model: Trains a linear SVM model on the scaled dataset.
4. Visualization: Visualizes the SVM decision boundary, maps routes, and displays feature importance.

Installation

Step 1: Clone the Repository

git clone https://github.com/your-username/pedestrian-friendly-cologne.git
cd pedestrian-friendly-cologne

Step 2: Install Dependencies

pip install -r requirements.txt

Usage

Run the script to generate the visualizations and results:

python pedestrian_analysis.py

The script fetches data from OpenStreetMap, simulates neighborhood data, trains an SVM model, and generates visualizations.

Visualizations

This project generates the following visualizations:

1. SVM Decision Boundary: Displays the classification of pedestrian-friendly areas.
2. Confusion Matrix: Evaluates the model's performance.
3. Feature Importance Plot: Highlights the most influential features in the SVM model.
4. Pedestrian Routes Map: Visualizes pedestrian, traffic, and cycleway routes in Cologne.

Results

The SVM model provides the following outputs:

• Classification Accuracy: Evaluated using a confusion matrix.
• Feature Importance: Determines which features most impact the pedestrian-friendly classification.
• Visual Map: Displays Cologne's pedestrian routes alongside traffic and cycleways.

Contributing

Contributions are welcome! If you have ideas for improvement or find any issues, feel free to:

1. Fork the repository.
2. Create a new branch for your feature or fix.
3. Submit a pull request.

License

This project is licensed under the MIT License.