Data set contains 29092 observation, out of which about 10% are default cases.

This model yield an AUC of 64%.

Method used - Logistic Regression

Next steps -

1. To see if the model improves using Decision Trees.
2. If decision trees does not suceed in improving, checking with random forest.