Customer Segmentation Data Warehouse Project

📌 Project Overview

The Customer Segmentation Data Warehouse (DW) project is designed to analyze customer spending behavior and segment customers into meaningful groups using data warehousing and machine learning techniques.

This project follows a complete ETL → Preprocessing → Dimensionality Reduction → Clustering → Prediction pipeline and finally allows the user to input spending amount and time (in days) to predict the customer segment.

---

🛠️ Technologies Used

• Python 3
• Pandas & NumPy – Data manipulation
• Scikit-learn – PCA & K-Means clustering
• Matplotlib / Seaborn – Data visualization
• Data Warehouse Concepts – ETL, analytical processing

---

📂 Project Structure

customer_segmentation_dw/
│
├── data/                   # Raw and processed datasets
├── scripts/                # Python scripts for each pipeline stage
│   ├── etl.py
│   ├── preprocessing.py
│   ├── pca.py
│   ├── clustering.py
│   └── user_segmentation.py
│
├── requirements.txt        # Required Python libraries
├── README.md               # Project documentation
└── .gitignore

---

⚙️ Setup Instructions

1️⃣ Install Python Dependencies

Make sure Python 3 is installed, then run:

pip install -r requirements.txt

---

▶️ How to Run the Project (Step-by-Step)

⚠️ Run the following commands one by one in the given order

---

🔹 Step 1: Run ETL Process

python3 scripts/etl.py

📌 This step:

• Extracts raw customer data
• Cleans and transforms it
• Loads it into a structured analytical format

---

🔹 Step 2: Data Preprocessing

python3 scripts/preprocessing.py

📌 This step:

• Handles missing values
• Applies feature scaling (standardization)
• Prepares data for PCA and clustering

---

🔹 Step 3: Apply PCA (Dimensionality Reduction)

python3 scripts/pca.py

📌 This step:

• Reduces high-dimensional data
• Preserves maximum variance
• Improves clustering performance and visualization

---

🔹 Step 4: Perform Clustering

python3 scripts/clustering.py

📊 Output:

• A customer clustering graph will appear on the screen

⚠️ Important: ➡️ Close the clustering graph window before proceeding further

---

🔹 Step 5: User-Based Customer Segmentation

python3 scripts/user_segmentation.py

📌 This step:

• Asks the user to input:

  • Amount spent
  • Time period (in days)

🧠 Based on the trained model, the system:

• Predicts the customer segment
• Displays the customer category on the terminal

---

📈 Example User Input

Enter amount spent: 12000
Enter time in days: 45

✅ Output:

Predicted Customer Segment: High-Value Customer

---

🧠 Conceptual Highlights (For Viva)

• Data Warehouse approach ensures structured analytical processing
• Standardization is used instead of normalization for PCA compatibility
• PCA reduces dimensionality while retaining variance
• K-Means clustering groups customers based on spending behavior
• User-driven prediction simulates real-world business decision making

---

🎯 Applications

• Customer behavior analysis
• Marketing strategy optimization
• Business intelligence systems
• Decision support systems

---

👨‍💻 Author

Shivam Gautam BCA – Data Warehousing & Analytics Project

---

📜 License

This project is for academic and learning purposes only.