Financial institutions usually use statistical models to estimate credibility of applicants for a loan. Applicant with the higest risk are declined, those who are appoved can be assigned different available product parameters based on their score.

Using provided data set, create an extreme gradient boosting model, that will serve for assessment of applicants creditbility. Quality of the model will be measured by its separation power - Gini coefficient.

To pass the competition sucessfully, competitiors should reach Gini higher than 52 Gini (equivalent to AUC 0.76) on the public leaderboard.