kunyuhe / ml-pipeline-on-financial-distress-data Goto Github PK

The project built a machine learning pipeline to classify whether an individual will experience financial distress (experience 90 days past due delinquency or worse) in the next two years based on income, household and credit history of that person. The pipeline currently supports 3 models (KNN, Decision Tree, and Random Forest), and use grid search and cross validation to find the best model in terms of either of the 2 metrics (Accuracy and Area Under the Receiver Operating Characteristic Curve).

License: GNU General Public License v3.0