K-Means Clustering with Multiple Distance Metrics

This Python script performs K-Means clustering on a given dataset using Euclidean, Cosine, and Jaccard distance metrics. The code includes data preprocessing, clustering implementation, SSE (Sum of Squared Errors) calculation, and accuracy assessment.

Requirements

• Python 3
• Pandas
• NumPy
• Scikit-learn

Usage

1. Data Preparation:

   • Ensure you have the required dataset (data.csv) and labels (label.csv) in the same directory as the script.

2. Run the Script:

   • Execute the Python script in a Python environment that has the necessary libraries installed.

   python kmeans_clustering.py

Code Overview

• Data Loading:

  • The script reads the dataset from data.csv using Pandas.

• Data Scaling:

  • The script scales the data to the range (0, 1).

• Random Centroids Initialization:

  • Random centroids are initialized for the K-Means algorithm.

• Clustering:

  • K-Means clustering is performed with Euclidean, Cosine, and Jaccard distance metrics.

• SSE Calculation:

  • SSE is calculated for each distance metric.

• Accuracy Assessment:

  • Majority voting is applied to assign labels, and accuracy is calculated for each distance metric.

Results

The script provides the SSE for each distance metric and the accuracy of K-Means clustering based on majority voting.

Notes

• This script assumes the availability of the required dataset and labels.
• Ensure that you have the necessary Python libraries installed.

Author

Vladislav Kuznetsov

License

This project is licensed under the MIT License.