Logistic regression, random forests, gradient boosting and XGboost are used. Feature permutation is used to validate further the model, as well as hyperparameter tuning. Everything is explained in the notebook, which comes with an associated library.
Like this project
0
Posted Jan 9, 2024
A credit score model created by testing several different deep learning methods. The result is an equation that gives the probability of not honouring a debt.