In previous videos we explained various scoring approaches: Expert scoring, Bayesian scoring and logistic regression scoring. The question pops-up on how to evaluate the performance of a scoring approach. We can do this by using a historical data set for which we know the loans that have defaulted.
The confusion matrix shows how many dossiers were evaluated correctly, and how many were displaced. This video we explain how to read and how to interpret the confusion matrix.

Links to other video's:
Expert scoring: [ Ссылка ]
Bayesian scoring: [ Ссылка ]
Logistic regression: [ Ссылка ]

свернуть