How astraldisaster11/credit-card-fraud-detection works
This is a Streamlit application for detecting credit-card fraud. Users upload training and test CSVs, explore the training data, preprocess and encode features, train selected classifiers, compare evaluation results, then use a trained model to predict on test data. The main workflow and capabilities are documented in the README and implemented in app.py. The excerpts show upload state, preprocessing, training, evaluation, and prediction; unsampled UI details are not expanded into additional components or links.
Main entry points: app.py Training Workflow
Components 16
Data Input
- CSV Upload
app.py - Session Datasets
app.py
Exploration and Preparation
- Data Exploration
app.py - Data Visualization
app.py - Data Cleaning
app.py - Feature Encoding
app.py
Model Training
- Train Test Split
app.py - Training Workflow
app.py - Logistic Regression
app.py - Random Forest
app.py - Decision Tree
app.py - Model Evaluation
app.py - Model Comparison
app.py
Prediction and Results
- Test Predictions
app.py - Prediction Results
app.py
Other components
- App User
Connections 20
- App User → CSV Upload: uploads CSVs
app.py - CSV Upload → Session Datasets: stores uploads
app.py - Session Datasets → Data Exploration: explores training data
app.py - Data Exploration → Data Visualization: renders charts
app.py - Data Exploration → Data Cleaning: prepares correlations
app.py - Data Cleaning → Feature Encoding: cleans features
app.py - Session Datasets → Training Workflow: provides training data
app.py - Training Workflow → Data Cleaning: cleans training data
app.py - Training Workflow → Feature Encoding: encodes training data
app.py - Training Workflow → Train Test Split: splits features
app.py - Training Workflow → Logistic Regression: trains selected model
app.py - Training Workflow → Random Forest: trains selected model
app.py - Training Workflow → Decision Tree: trains selected model
app.py - Training Workflow → Model Evaluation: evaluates predictions
app.py - Model Evaluation → Model Comparison: compares accuracy
app.py - Session Datasets → Test Predictions: provides test data
app.py - Test Predictions → Data Cleaning: prepares test data
app.py - Test Predictions → Feature Encoding: encodes test data
app.py - Test Predictions → Random Forest: uses trained model
app.py - Test Predictions → Prediction Results: shows predictions
app.py