Assignment 4: Customer Segmentation with Clustering

Overview

This project implements a customer segmentation analysis for a UK-based online gift retailer using K-means clustering. The analysis segments customers based on their transaction history (2009-2011) using RFM (Recency, Frequency, Monetary) features to identify distinct customer groups and provide targeted marketing recommendations.

Key Objectives

• Aggregate transaction data to customer-level RFM features
• Determine optimal number of clusters using elbow method and silhouette analysis
• Train K-means clustering model to segment customers
• Visualize and interpret customer segments
• Provide actionable business recommendations for each segment

Project Structure

Assignment-4/
├── README.md                    # Project documentation (this file)
├── starter_notebook.ipynb       # Main Jupyter notebook with analysis
└── data/
    └── online_retail_II.csv     # Transaction dataset (2009-2011)

Data Description

The online_retail_II.csv dataset contains transaction records with the following key fields:

• Invoice: Unique transaction identifier
• StockCode: Product code
• Description: Product name
• Quantity: Number of items purchased
• InvoiceDate: Transaction date and time
• Price: Unit price per item
• Customer ID: Unique customer identifier
• Country: Customer's country

How to Run

Option 1: Using Jupyter Notebook in VS Code

1. Open the notebook:

   • Open starter_notebook.ipynb in VS Code

2. Install dependencies:

   • Run the first code cell to install required packages:

     pip install pandas matplotlib seaborn scikit-learn

3. Run the analysis:

   • Execute cells sequentially from top to bottom
   • Fill in the TODO sections as you progress through the assignment

Option 2: Using Terminal

1. Install dependencies:

   pip install pandas matplotlib seaborn scikit-learn

2. Launch Jupyter:

   jupyter notebook starter_notebook.ipynb

   Or:

   jupyter lab

3. Execute the notebook:

   • Run cells in order, completing the TODO sections

Notebook Structure

The analysis is organized into six main steps:

Step 1: Import Libraries and Load Data

• Import required Python libraries
• Load the Online Retail II dataset
• Display basic dataset information

Step 2: Aggregate Transaction Data to Customer-Level RFM Features

• Clean transaction data (remove missing IDs, returns, etc.)
• Calculate RFM metrics for each customer:
  • Recency: Days since last purchase
  • Frequency: Number of unique transactions
  • Monetary: Total amount spent

Step 3: Standardize Features and Determine Optimal k

• Standardize RFM features using StandardScaler
• Apply elbow method to test k values from 2 to 10
• Calculate silhouette scores for cluster quality validation
• Select optimal k value with written justification

Step 4: Train K-Means Model and Visualize Segments

• Train final K-means model with optimal k
• Add cluster labels to customer data
• Create scatter plot visualization (Frequency vs Monetary)
• Calculate and display cluster centers

Step 5: Interpret Segments and Provide Business Recommendations

• Analyze each customer segment's characteristics
• Provide descriptive names for segments
• Calculate detailed statistics for each segment

Step 6: Submit Your Work

• Verify all code runs without errors
• Push to GitHub
• Submit repository link

Submission

After completing all steps:

git add .
git commit -m 'completed customer segmentation assignment'
git push

Submit your GitHub repository link on the course platform.