Customer Segmentation using K-Means Clustering

An unsupervised machine learning project that groups retail customers according to their annual income and spending score. The resulting segments can support targeted marketing, customer retention, and purchasing-pattern analysis.

---

📋 Table of Contents

• Overview
• Project Objective
• Technologies Used
• Methodology
• Dataset
• Project Structure
• Requirements
• Setup
• How to Run
• Results
• Customer Segments
• Learning Outcomes
• Future Improvements
• Author
• Internship Details

---

🎯 Overview

This project is part of the SkillCraft Technology Machine Learning Internship and uses K-Means clustering to identify groups of customers with similar income and spending behavior.

Workflow:

• Load and inspect the Mall Customers dataset
• Select annual income and spending score as clustering features
• Use the Elbow Method to estimate the number of clusters
• Train a K-Means model with five clusters
• Visualize clusters and their centroids
• Export the clustered customer data to CSV

---

🎯 Project Objective

Segment customers into meaningful groups using Annual Income (k$) and Spending Score (1-100), then interpret those groups for business decision-making.

---

🛠 Technologies Used

• Python 3
• Pandas
• NumPy
• Scikit-learn
• Matplotlib
• K-Means clustering

---

🔬 Methodology

1. Load the customer data from data/Mall_Customers.csv.
2. Select the income and spending-score columns.
3. Calculate WCSS for values of k from 1 to 10.
4. Apply the Elbow Method to select the cluster count.
5. Fit K-Means with random_state=42 and n_init=10.
6. Label clusters using their centroid income and spending levels.
7. Save a combined Elbow Method and segmentation plot.
8. Export the results with cluster labels to a CSV file.

---

📊 Dataset

The project uses the Mall Customers dataset. The model uses these two features:

Feature	Description
Annual Income (k$)	Customer annual income in thousands of dollars
Spending Score (1-100)	Customer spending score

The dataset contains 200 customer records. If the local CSV is unavailable, the script creates a reproducible sample dataset so it can still be executed.

---

📁 Project Structure

SCT_ML_2/
├── README.md
├── kmeans_clustering.py
├── data/
│   └── Mall_Customers.csv
├── customer_segments_plot.png
└── segmented_customers.csv

---

📦 Requirements

Python 3.9 or later is recommended. Install the dependencies with:

pip install pandas numpy matplotlib scikit-learn

---

⚙️ Setup

1. Open the project folder

cd "SCT_ML 2"

2. Create and activate a virtual environment (optional)

Windows PowerShell:

python -m venv .venv
.\.venv\Scripts\Activate.ps1

Windows Command Prompt:

python -m venv .venv
.venv\Scripts\activate.bat

3. Install dependencies

pip install pandas numpy matplotlib scikit-learn

---

▶️ How to Run

python kmeans_clustering.py

The script:

• prints dataset and cluster information
• saves customer_segments_plot.png
• saves clustered data to segmented_customers.csv

---

📈 Results

The model divides the customers into five clusters and produces a plot containing both the Elbow Method curve and the customer segments. Each point represents a customer, while the marked centroids represent the center of each cluster.

---

👥 Customer Segments

• Target: High income, high spending customers
• Careful: High income, low spending customers
• Standard: Moderate income and moderate spending customers
• Careless: Low income, high spending customers
• Sensible: Low income, low spending customers

---

🎓 Learning Outcomes

• Understanding unsupervised learning and K-Means clustering
• Selecting the number of clusters with WCSS and the Elbow Method
• Applying feature selection to a distance-based algorithm
• Interpreting cluster centroids
• Visualizing customer groups and exporting predictions

---

🚀 Future Improvements

• Scale features before clustering and compare the results
• Evaluate the solution using silhouette score
• Add demographic features such as age and gender
• Build an interactive dashboard for segment exploration

---

👨‍💻 Author

Devarakonda Vignesh Varsha

---

📌 Internship Details

Detail	Information
Internship	SkillCraft Technology
Domain	Machine Learning
Task	Task 02 - Customer Segmentation using K-Means Clustering
Clusters	5

---

Built with ❤️ during SkillCraft Technology ML Internship