You can deploy this model and start earning money today!

deathlegionteam/
fraud-detection-model

Name: fraud-detection-model
Availability: InStock

joblib

Views

No views yet

🔥 Death Legion Fraud Detection System

Death Legion

Best Teams

🎯 Overview

Welcome to the Death Legion Fraud Detection System - a state-of-the-art machine learning solution for real-time credit card fraud detection. Built by the Best Teams with cutting-edge Random Forest technology, this model achieves exceptional performance on highly imbalanced financial datasets.

🚀 Deploy This Space

🎮 Use This Model

⚡ Key Features

🛡️ Advanced Random Forest Architecture: 500 estimators with optimized depth
📊 Superior Performance: AUPRC 0.8177 on imbalanced fraud data
🔒 Secure Safetensors Format: Fast, safe model serialization
⚡ Real-time Inference: Sub-millisecond prediction latency
🎯 Imbalanced Data Optimized: Precision-Recall focused evaluation

📈 Performance Metrics

Metric	Score	Status
AUPRC	0.8177	✅ Excellent
Precision	0.8182	✅ High
Recall	0.8265	✅ Strong
F1-Score	0.8223	✅ Balanced

🏗️ Model Architecture

Random Forest Classifier
├── Estimators: 500
├── Max Depth: 25
├── Features: 30 (V1-V28 + Time + Amount)
├── Classes: 2 (Legitimate, Fraudulent)
└── Format: Safetensors (6.12 MB)

📖 How to Use

Quick Start

Access the Live Demo: Visit https://huggingface.co/spaces/Pnny13/fraud-detection-space
Enter Transaction Features: Input the 30 features (V1-V28, Time, Amount) from your credit card transaction
Get Instant Prediction: The system will return:
- Fraud probability score (0-100%)
- Binary classification (Fraud/Legitimate)
- Recommendation for action

Programmatic Usage

python

from safetensors.numpy import load_file
import numpy as np

# Load model from Hugging Face Hub
from huggingface_hub import hf_hub_download
model_path = hf_hub_download(
    repo_id="Pnny13/fraud-detection-model",
    filename="fraud_detector.safetensors"
)

# Load and predict
tensors = load_file(model_path)
# ... use model for predictions

📁 Repository Structure

Pnny13/fraud-detection-space/
├── app.py              # Gradio application
├── requirements.txt    # Python dependencies
└── README.md           # Documentation

Pnny13/fraud-detection-model/
├── fraud_detector.safetensors  # Trained model
├── scaler.joblib               # Feature scaler
├── inference.py                # Prediction script
└── README.md                   # Model documentation

🔬 Technical Details

Dataset

Source: Kaggle - Credit Card Fraud Detection
Transactions: 284,807 total
Fraud Cases: 492 (0.172% - highly imbalanced)
Features: 30 PCA-transformed features + Time + Amount

Preprocessing

Robust Scaling: Applied to Time and Amount features
Feature Engineering: PCA components V1-V28
Class Balancing: Optimized for precision-recall tradeoff

Training Configuration

python

RandomForestClassifier(
    n_estimators=500,
    max_depth=25,
    class_weight='balanced_subsample',
    random_state=42,
    n_jobs=-1
)

🎮 Live Demo

Try the live interactive demo at: https://huggingface.co/spaces/Pnny13/fraud-detection-space

🤝 Credits

Powered by Death Legion
Elite Machine Learning Division

Best Teams Collaboration
Excellence in AI Engineering

📜 License

This model is released under the MIT License. Use responsibly for fraud detection and financial security applications.

🔗 Links

🤗 Model: https://huggingface.co/Pnny13/fraud-detection-model
🚀 Space: https://huggingface.co/spaces/Pnny13/fraud-detection-space
📊 Dataset: https://www.kaggle.com/datasets/mlg-ulb/creditcardfraud

_{Built with 🔥 by Death Legion | Best Teams Elite Division}

deathlegionteam/fraud-detection-model