Capital One Data Scientist interview questions

Case Interview: the case is the car finance loan. - what are revenues and expenses - given a model that predicts when a customer is good (loan should be approved) or bad (loadn should be decline) find out: 1. the probability that the customer is good given the model predicts good 2. the probability that the customer is bad given the model is good 3. given a pentile graph of # of checked off loans / # of loans what is a better model than the current; what is the best model. Behavioral interview: - tell me about a time that you had to deal with changing objectives in your team/project - tell me about a time that you had to deal with unexpected problems in your project - tell me about a time that you had to persuase somebody Role interview: the case is a report on air company with low percentage of flight on time. Read the report an give an evaluation of it and some reccomendations to your boss. 15 minutes to read the report and remove anything unecessary or spot errors. 20 minutes to present it to your boss. 15 minutes to discuss afterwards from data scientist to data scientist.

Questions were based on the data set which they had given. (Analysis stuff mostly)