https://github.com/preethi2805/customer-segmentation

This project applies Recency, Frequency, and Monetary (RFM) Analysis along with K-Means Clustering to segment customers based on their purchasing behavior. The goal is to identify distinct customer groups and develop targeted marketing strategies.
https://github.com/preethi2805/customer-segmentation

customer-segmentation kmeans-clustering python-3 rfm-analysis

Last synced: 24 days ago
JSON representation

This project applies Recency, Frequency, and Monetary (RFM) Analysis along with K-Means Clustering to segment customers based on their purchasing behavior. The goal is to identify distinct customer groups and develop targeted marketing strategies.

• Host: GitHub
• URL: https://github.com/preethi2805/customer-segmentation
• Owner: Preethi2805
• Created: 2025-01-28T04:50:43.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-02-20T05:27:31.000Z (over 1 year ago)
• Last Synced: 2025-02-20T05:29:56.688Z (over 1 year ago)
• Topics: customer-segmentation, kmeans-clustering, python-3, rfm-analysis
• Language: Jupyter Notebook
• Homepage:
• Size: 22.6 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# 📊 Customer Segmentation using RFM and K-Means Clustering

## 🔍 Problem Statement

Understanding customer behavior is crucial for businesses. By segmenting customers based on their purchasing patterns, businesses can:

- Identify **high-value** customers.

- Recognize **at-risk** customers.

- Personalize marketing strategies for **customer retention** and **acquisition**.

## 📌 Dataset

The dataset consists of transaction records from an e-commerce platform with the following key attributes:

- `InvoiceNo` – Unique invoice identifier.

- `StockCode` – Product identifier.

- `Description` – Product description.

- `Quantity` – Number of units purchased.

- `InvoiceDate` – Date of purchase.

- `UnitPrice` – Price per unit.

- `CustomerID` – Unique customer identifier.

- `Country` – Country of purchase.

## 🛠️ Methodology

### 1️⃣ Data Preprocessing

- Handled missing values and duplicates.

- Converted `InvoiceDate` to a datetime format.

- Removed outliers using **Z-score analysis**.

### 2️⃣ RFM Analysis

- **Recency (R)**: Days since the customer’s last purchase.

- **Frequency (F)**: Number of transactions by each customer.

- **Monetary Value (M)**: Total spend per customer.

### 3️⃣ K-Means Clustering

- Determined the optimal number of clusters using the **Elbow Method** and **Silhouette Score**.

- Segmented customers into **4 clusters**.

### 4️⃣ Insights from Clusters

- **Cluster 3:** Loyal and high-spending customers.

- **Cluster 2:** Less engaged customers with high recency.

- **Cluster 1 & 0:** Intermediate customers who can be nurtured.

## 📈 Results & Visualizations

- **Pair plots** to analyze feature distributions across clusters.

- **Box plots** to compare `Recency`, `Frequency`, and `Monetary Value` across clusters.

- **Bar charts** showing average RFM values per cluster.

![Pairplot - relationships between features and their cluster distribution.](pairplot.png)

![Boxplot](boxplot.png)

## 🔧 Technologies Used

- **Python**

- **Pandas, NumPy** – Data Manipulation

- **Matplotlib, Seaborn** – Data Visualization

- **Scikit-Learn** – Machine Learning (K-Means Clustering)

- **StandardScaler** – Feature Scaling

## 📌 Future Enhancements

- Implement **Hierarchical Clustering** for better interpretability.

- Develop **automated customer insights** with dashboards.

- Integrate with **real-time e-commerce data**.