You can deploy this model and start earning money today!

JosianaSilva/
case-deteccao-de-fraude

Name: case-deteccao-de-fraude
Availability: InStock

Views

No views yet

Fraud Detection Model

Credit Card Fraud Detection Model using Logistic Regression.

Metrics

Accuracy: 0.9509 (95.09%)
Roc Auc: 0.8342 (83.42%)
Precision: 0.9089
Recall: 0.6783
F1 Score: 0.7768

Deployment Criteria

This model was automatically deployed because it meets the following criteria:

Accuracy > 90%
ROC AUC Score > 75%

Usage

python

from huggingface import load_model_from_hf
import pandas as pd

model, feature_names, scaler = load_model_from_hf("your-repo-id")

data = pd.DataFrame(...)  
X_scaled = scaler.transform(data[feature_names])

predictions = model.predict(X_scaled)
probabilities = model.predict_proba(X_scaled)[:, 1]

Model Architecture

Algorithm: Logistic Regression (scikit-learn)
Features: Preprocessed transaction data with categorical encoding
Preprocessing: StandardScaler normalization

Training Details

Data Processing:
- Temporal features: hour, day_of_week, month
- Age calculation from date of birth
- One-Hot Encoding: category, state
- Target Encoding: merchant, city, job
Train/Test Split: 80/20 with random_state=42
Model Parameters: max_iter=1000, random_state=42

Files Included

model.pkl: Trained Logistic Regression model
scaler.pkl: StandardScaler for feature normalization
feature_names.pkl: List of feature names for proper data alignment

JosianaSilva/case-deteccao-de-fraude