Credit_Card_Fraud_Synthetic_Detector

💳 Credit Card Fraud & Anomaly Detection System

An interactive, production-ready Machine Learning Web Application built with Streamlit to detect fraudulent credit card transactions in real-time based on transformed PCA features and transaction metrics.

---

🖼️ Application Preview & UI

Interactive Streamlit dashboard designed with business-friendly feature inputs, fault-tolerant logic, and instant fraud probability metrics.

---

🔗 Live Application

You can access and test the deployed application directly here: 👉 Click Here to Launch Live Demo

---

🖼️ Key Features & Capabilities

• User-Friendly Interface: Business-oriented feature labels replacing raw mathematical PCA names ($V1$–$V5$) for seamless user experience.
• Smart Decision Dashboard: Clear status classification (Legitimate vs. Fraudulent), Risk Levels, and Model Confidence Score ($0-100%$) instead of raw binary values ($0/1$).
• Fault-Tolerant Pipeline: Robust exception handling ensuring that the UI renders smoothly even if dependencies or .pkl files are missing.
• High Accuracy Benchmarks: Powered by high-precision ensemble tree classifiers trained on transaction risk patterns.

---

📊 Model Evaluation & Benchmarks

Multiple classification algorithms were evaluated to handle high variance and anomaly patterns. Here is the comparative accuracy metric across models:

Selected Baseline: The Random Forest Classifier achieved 98.0% accuracy with reliable probability estimation for risk grading.

---

⚙️ Model Features & Input Parameters

---

📁 Repository Structure

├── credit_card_fraud_synthetic.csv     # Primary Transaction Dataset
├── model.ipynb                         # Data Preprocessing, EDA & Model Training
├── .gitignore                          # Git Ignore Rules
├── .gitattributes                      # Git LFS Configuration
├── app.py                              # Fault-Tolerant Streamlit Web Application
├── model.pkl                           # Trained Machine Learning Model
├── scaler.pkl                          # Fitted StandardScaler Object
├── column.pkl                          # Feature Column Definitions
├── requirements.txt                    # Python Dependencies
└── README.md                           # Project Documentation

git clone https://github.com/amirsohail100/Credit-Card-Fraud-Detection-System.git

cd Credit-Card-Fraud-Detection-System

streamlit run app.py

pip install -r requirements.txt

📄 License

This project is licensed under the MIT License.

📝 Author

👤 Amir Sohail

A high-accuracy Machine Learning project evaluating multiple classifiers (Random Forest, Extra Trees, XGBoost, KNN at 98% accuracy) on Credit Card Fraud Data. Features dynamic inputs like Time, PCA components (V1-V5), and Amount to predict transaction legitimacy (Class) via a fault-tolerant Streamlit web UI.