clustering-eval 0.1.1

A package for evaluating clustering algorithms using NCQI.

Clustering Evaluation (NCQI)

A Python package for evaluating clustering algorithms using the Normalized Clustering Quality Index (NCQI). This package is designed for researchers and practitioners working with unsupervised clustering techniques, providing a quantitative measure for cluster quality assessment.

📌 Installation

To install the package, run:

pip install git+https://github.com/Mojtaba-jahanian/Cosine-Clustering-Index-CCI-.git

📖 Reference Paper

This implementation is based on the research paper: 🔗 "Cosine Clustering Index (CCI) for Deep Clustering Evaluation"

📌 Authors: Mojtaba Jahanian, [Other Authors]

🚀 Usage

from clustering_eval.ncqi import normalized_clustering_quality_index
import numpy as np

# Generate random data and labels
X = np.random.rand(100, 10)
labels = np.random.randint(0, 3, size=100)

# Compute NCQI Score
ncqi_score = normalized_clustering_quality_index(X, labels)
print("NCQI Score:", ncqi_score)

🏆 Features

• Evaluates clustering quality using a novel metric based on cohesion and separation.
• Supports multiple clustering algorithms including KMeans, Agglomerative Clustering, DBSCAN, and Spectral Clustering.
• Scalable for large datasets such as CIFAR-10.
• Easy integration into machine learning workflows.

📊 Example: Clustering CIFAR-10

python examples/cifar10_clustering.py

This example performs clustering on the CIFAR-10 dataset and evaluates the clustering results using NCQI.

🔹 Creating and Publishing the Package

Step 1: Build the Package

python setup.py sdist bdist_wheel

✅ This command creates dist/ and build/ folders containing the final package files.

Step 2: Upload to PyPI

twine upload dist/*

🔹 Enter your PyPI username and password when prompted.

Step 3: Install from PyPI

pip install clustering_eval

📜 License

This project is licensed under the MIT License.

📥 Download the package: clustering_eval_package.zip