How jaymonty/credit-risk-analysis works
This repository documents a notebook-based credit-risk modeling project: credit-related data is explored and prepared, several classifiers are trained with and without SMOTE and RFECV, and their results are compared to identify a tuned model. The README also names deployment as a project objective, but the tree shows no deployment artifact, so it is represented only as an undocumented implementation. Relationships below follow the README’s stated workflow; notebook contents were not available for inspection.
Main entry points: Model_Dev_.ipynb Classifiers · Data_Preprocessing.ipynb Data exploration
Components 12
Data preparation
- Data exploration
Data_Preprocessing.ipynb - Data cleaning
Data_Preprocessing.ipynb - Data preprocessing
Data_Preprocessing.ipynb
Model development
- Classifiers
Model_Dev_.ipynb - SMOTE sampling
Model_Dev_.ipynb - RFECV selection
Model_Dev_.ipynb - Hyperparameter tuning
Model_Dev_.ipynb
Evaluation and use
- Model evaluation
Model_Dev_.ipynb - Model comparison
Model_Dev_.ipynb - Deployment
Other components
- Financial institution
- Credit records
Connections 13
- Credit records → Data exploration: explores
README.md - Data exploration → Data cleaning: informs
README.md - Data cleaning → Data preprocessing: prepares data
README.md - Data preprocessing → Classifiers: supplies data
README.md - Classifiers → SMOTE sampling: retrained with
README.md - Classifiers → RFECV selection: retrained with
README.md - Classifiers → Hyperparameter tuning: tunes best model
README.md - Classifiers → Model evaluation: evaluated by
README.md - SMOTE sampling → Model evaluation: evaluated by
README.md - RFECV selection → Model evaluation: evaluated by
README.md - Model evaluation → Model comparison: compared in
README.md - Model comparison → Financial institution: supports decisions
README.md - Model comparison → Deployment: intended for
README.md