This project aims to detect fraudulent credit card transactions using machine learning algorithms. The system incorporates various features such as biometric authentication, data visualizations, synthetic data testing, location-based fraud detection, and secure databases to ensure robust fraud detection capabilities.
-
Biometric Authentication
- Ensures secure access to the fraud detection page.
-
Data Visualizations
- Confusion Matrix: Displays true positives, false positives, true negatives, and false negatives.
- Bar Plots: Shows the distribution of fraudulent and non-fraudulent transactions.
- Line Charts: Tracks transaction trends and anomalies over time.
-
Model Accuracy
- The machine learning model achieves an accuracy of 0.99, indicating high reliability in detecting fraudulent transactions.
-
Synthetic Data Testing
- Synthetic data with 10 transactions is generated for testing. Normal transactions are highlighted in green, and fraudulent transactions in red. Transaction IDs detected as fraud are also displayed.
-
Location-Based Fraud Detection
- Detects fraud based on the user's current location. Any transaction initiated from a location other than Kolkata is flagged as potentially fraudulent.
-
Company-Based Fraud Detection
- Uses a dataset of company names commonly reported for involvement in scams. Transactions associated with companies from this dataset are considered fraudulent.
-
Real-Time Alerts
- If fraud is detected, an alert message is sent to the user's registered phone number with relevant transaction details.
-
Multiple Transactions Detection
- Detects fraud when multiple transactions of the same amount (less than the amount required for OTP) are made back-to-back.
-
Secure Database
- Ensures that all transaction data is stored securely.