💳 Credit Card Fraud Detection — ML Classification Pipeline

An end-to-end Machine Learning project for detecting fraudulent credit card transactions using ensemble learning techniques. This project benchmarks multiple tree-based models and demonstrates superior performance using gradient boosting methods on a highly imbalanced dataset.

Credit card fraud detection is a highly imbalanced binary classification problem where fraudulent transactions represent a very small percentage of total transactions.

The objective of this project is to:

• Build a robust fraud detection pipeline
• Compare multiple ensemble learning models
• Evaluate performance using ROC-AUC and classification metrics
• Identify the best performing model for real-world deployment

---

📊 Dataset Information

• Real-world credit card transaction dataset

• Highly imbalanced (~0.17% fraud cases)

• Features:

  • PCA transformed features (V1–V28)
  • Time
  • Amount
  • Target variable Class (0 = Legitimate, 1 = Fraud)

• Dataset Link : https://www.kaggle.com/datasets/mlg-ulb/creditcardfraud

---

⚙️ Tech Stack

• Python
• NumPy
• Pandas
• Scikit-learn
• XGBoost
• LightGBM
• CatBoost
• Matplotlib
• Seaborn

---

🧠 Models Implemented

The following ensemble models were trained and evaluated:

• Random Forest
• AdaBoost
• CatBoost
• LightGBM
• XGBoost

---

📈 Model Performance (ROC-AUC Score)

Model	ROC-AUC
🥇 XGBoost	0.9771
🥈 LightGBM	0.9682
CatBoost	0.8578
Random Forest	0.8529
AdaBoost	0.8135

🔥 Best Model: XGBoost

• Achieved ROC-AUC = 0.977
• Excellent class separation capability
• Strong performance on imbalanced dataset
• Suitable for production-level fraud detection systems

---

📊 Evaluation Metrics Used

• ROC Curve & AUC Score
• Confusion Matrix
• Precision
• Recall
• F1-Score

Special focus was given to Recall to minimize false negatives (missed fraud cases).

---

🗂 Project Structure

Fraud_Detection_Model/
│
├── credit_card_fraud_detection.ipynb
├── README.md
└── requirements.txt

---

🚀 How to Run the Project

1️⃣ Clone the repository

git clone https://github.com/ShrutiPatel263/Fraud_Detection_Model.git

2️⃣ Install dependencies

pip install -r requirements.txt

3️⃣ Run the notebook

jupyter notebook credit_card_fraud_detection.ipynb

---

📌 Key Highlights

✔ Implemented complete ML pipeline ✔ Compared 5 ensemble learning algorithms ✔ Handled severe class imbalance ✔ Achieved high ROC-AUC (0.977) ✔ Conducted systematic model benchmarking

---

👩‍💻 Author

Shruti Patel Machine Learning & AI Enthusiast

GitHub: https://github.com/ShrutiPatel263

---