Feature description

I propose adding an implementation for the K-Medoids clustering algorithm to this repository.

• K-Medoids is a classic clustering technique, similar to K-Means, but uses actual data points (medoids) as cluster centers, making it more robust to noise and outliers.
• Useful for comparing clustering algorithms or datasets where K-Means does not perform well.
• Should include clear implementation, usage examples, and validation tests.