Student Clustering System using Python + Machine Learning (on CGPA)

📌 Project Summary

The Student Clustering System is a web-based application developed using Python and Machine Learning that groups students based on their CGPA and other numerical data. Built with the Streamlit framework, this system allows users to upload student data in CSV format and perform KMeans clustering. The results are presented in tabular form, along with multiple data visualizations, and can be downloaded directly.

Download New Real Time Projects :-Click here

🔧 Technologies Used

• Python
• Streamlit – Web Interface
• Pandas – Data Manipulation
• scikit-learn – For KMeans Clustering & Data Scaling
• Seaborn / Matplotlib – Data Visualizations

🔄 Project Flow: Step-by-Step

1. Upload CSV File

• User uploads a .csv file.
• The system previews the data and filters only numeric columns for clustering.

2. Configure Clustering

• A slider allows the user to select the number of clusters (K).
• Data is standardized using StandardScaler.
• KMeans algorithm clusters the data.
• A new column Cluster is added to the dataset showing cluster labels.

3. Show Clustered Data

• The table with clusters is displayed.
• Users can download the clustered data as a CSV file.

4. Visualizations (Charts)

The system generates four types of visualizations to interpret the clusters:

• ✅ Bar Chart – Cluster Sizes: Shows the number of students in each cluster.
• ✅ Scatter Plot: Plots the first two numeric features, color-coded by cluster.
• ✅ Pie Chart: Represents the distribution of a selected numeric feature across clusters.

5. Cluster Info (Descriptions)

Interpretations based on CGPA:

• Cluster 0: Low CGPA (below 6.5)
• Cluster 1: Average CGPA (6.5 – 8.0)
• Cluster 2: High CGPA (above 8.0)

🎯 Key Features

• ✅ Upload any student-related CSV with numeric features
• ✅ Automatically selects numeric columns
• ✅ KMeans clustering with customizable number of clusters
• ✅ Four interactive charts to visualize results
• ✅ Downloadable clustered data
• ✅ Works with all types of academic datasets (not limited to CGPA)