Data Analysis and Clustering Application

This project, developed in Python with Jupyter and Tkinter, is a complete application for data exploration and the application of various clustering algorithms. It was carried out as part of the Data Mining module (M1 Bioinformatics, USTHB). The application offers an intuitive graphical interface allowing to carry out a complete data mining pipeline, from preprocessing to cluster performance evaluation.

📸 Screenshots

Main application interface offering preprocessing and clustering tools. Example of the Elbow method to determine the optimal number of clusters.

✨ Features

1. Preprocessing and Exploratory Data Analysis (EDA)

• Data Loading: Opening .csv and .arff files.
• Exploration: Display of distinct values, number of instances, attribute information.
• Descriptive Statistics: Calculation of Min, Max, Mean, Median and Quartiles.
• Visualization: Generation of Boxplots and Scatter plots to analyze distributions and correlations.
• Cleaning: Detection and replacement of missing values.
• Normalization: Data scaling via Min/Max and Z-score methods.

2. Clustering Algorithms

The application implements and compares 5 major clustering algorithms:

• K-Means: Based on centroids.
• K-Medoids: Robust variant using medoids.
• AGNES (Agglomerative Nesting): Hierarchical agglomerative clustering.
• DIANA (Divisive Analysis): Hierarchical divisive clustering.
• DBSCAN (Density-Based): Density-based clustering, capable of finding arbitrary shapes.

3. Performance Evaluation

• Elbow Method: To determine the optimal number of clusters (k).
• Inertia Measures: Calculation of within-cluster and between-cluster inertia.
• Clustering Scores: Evaluation via Silhouette Score, Calinski-Harabasz Index and Davies-Bouldin Index.

🛠️ Technologies and Libraries

• Python 3
• Tkinter for the graphical interface (GUI)
• Jupyter Notebook for the development environment
• Pandas & Numpy for data manipulation
• Scikit-learn & Scikit-learn-extra for clustering algorithms and preprocessing
• Matplotlib & Seaborn for data visualization
• Kneed for Elbow curve analysis

🚀 How to Use

1. Clone the repository:

   git clone https://github.com/YOUR_USERNAME/Data-Clustering-Toolkit.git
   cd Data-Clustering-Toolkit

2. Install dependencies:

   pip install -r requirements.txt

3. Launch Jupyter Notebook:

   jupyter notebook

4. In the Jupyter interface, navigate to the notebook/ folder and open Clustering_Analysis.ipynb.
5. Execute the notebook cells. The last cell will launch the Tkinter graphical application.

📌 Citation

If you use this project, please cite it as:

Ayoub Laib (2025), Data Analysis and Clustering Application, GitHub repository: https://github.com/aylaib/Data-Clustering-Toolkit

📚 Reference Documents

• Complete Project Report : Detailed analysis, methodology, results and conclusions.