tanyakuznetsova / Multidimensional-Scaling-of-European_Cities

This project explores the spatial relationships between twenty European cities using classical manual Multidimensional Scaling (MDS), MDS from scikit-learn, and compares the results with Principal Component Analysis (PCA).

Multidimensional-Scaling-of-European_Cities

This project explores the spatial relationships between twenty European cities using classical manual Multidimensional Scaling (MDS), MDS from scikit-learn, and compares the results with Principal Component Analysis. It contains the visualizations and analysis results.

Contents

• Multidimensional scaling of European Cities.ipynb: The Jupyter notebook containing the project.
• README.md: Brief overview of the project.

Execution Environment

The project is developed and executed in a Google Colab environment.

Execution Flow

1. Classical manual MDS: Implementing Multidimensional Scaling manually to explore spatial relationships.
2. MDS from scikit-learn: Utilizing the MDS implementation from scikit-learn library for comparison.
3. PCA Comparison: Applying Principal Component Analysis to compare results with MDS.

Dependencies

• Python
• NumPy
• scikit-learn
• Matplotlib
• PCA

Note

• The project discusses the spatial relationships between cities using dimensionality reduction techniques.

• MDS may involve manual rotation or mirroring to achieve desired visualizations.

• PCA might also result in rotated or mirrored visualizations depending on the orientation of principal components.

• Acknowledgements

• The distances between cities were obtained from https://www.distancecalculator.net/