📊 Credit Card Fraud Detection using SQL

📖 Overview

This project analyzes credit card transaction data to detect fraudulent activities using rule-based methods in SQL.

📊 Dataset

• Source: Credit Card Transactions Dataset

  https://www.kaggle.com/datasets/priyamchoksi/credit-card-transactions-dataset?select=credit_card_transactions.csv

• Size used: Sample 50,000 transactions

• Includes customer ID, transaction, amount, location, and fraud label

🔍 Key Analysis

Fraud Distribution

• Fraud accounts for 0.9% of all transactions
• Highly imbalanced dataset

💳 Transaction Behavior

• Fraudulent transactions have significantly higher values
• Average fraud transaction is ~8x larger than normal

⚙️ Detection Rules

Rule 1: Amount Anomaly

• Flag transactions > 3x customer average
• Recall ≈ 46%
• Precision ≈ 9%

Rule 2: Rapid Transactions

• Flag transactions within short time intervals
• Minimal impact on fraud detection

Combined Rules

• Recall ≈ 47%
• Precision ≈ 9%

💡 Key Insights

• Fraud is strongly associated with unusually high transaction amounts
• Transaction frequency is not a strong fraud indicator in this dataset
• Rule-based detection produces many false positives

🚀 Business Recommendations

• Monitor high-value transactions in real-time
• Combine multiple behavioral signals
• Implement machine learning models for improved accuracy

🛠️ Tools Used

• SQL (MySQL)
• Python (Pandas)

📁 Project Structure

├── data/ ├── sql/ │ └── fraud_detection.sql ├── README.md