feat: implement K-Means and Hierarchical clustering with evaluation o…

[codermiki]

Overview

This pull request introduces an unsupervised machine learning analysis using the wine quality (red and white) dataset. The work focuses on clustering wines based on physicochemical properties without using quality labels during training.

Key Changes

• Added full data preprocessing and feature scaling workflow
• Implemented K-Means clustering with optimal K selection using the Elbow Method
• Evaluated K-Means clustering using Silhouette Score
• Implemented Hierarchical Clustering with dendrogram visualization
• Provided post-cluster interpretation using wine quality for validation

Algorithms Used

• K-Means Clustering
• Hierarchical Clustering (Ward linkage)

Outcome

The clustering results show meaningful separation of wines based on chemical composition, with clusters exhibiting distinct average quality levels, validating the effectiveness of the unsupervised approach.

Notes

• The quality column was excluded from training and used only for post-analysis
• All features were standardized to ensure fair distance calculations