German-Credit-Risk-Classification

Machine Learning Classification with german credit data from UCI Machine Learning Repository: https://archive.ics.uci.edu/ml/datasets/Statlog+(German+Credit+Data)

Files

• File germancredit contains data visalisation, preprocessing steps and literally all that needed to be done in order to find the best model incl. parameter settings.
• File Final_Model contains the final, best classifying models

Applied Algorithms with python scikit-learn:

• SVC
• Gaussian Naive Bayes
• Randomforest Classifier
• Extratrees Classifier
• Gradient Boosting Classifier
• AdaBoost Classifier
• Bagging Classifier

Evaluation

Best Algorithm is Gradient Boosting Classifier with a 10-fold Cross-Validation:

• Cross Validation Precision: 0.85 (+/- 0.17)
• Cross Validation Recall: 0.86 (+/- 0.04)
• Cross Validation roc_auc: 0.91 (+/- 0.09)