Peer-graded Assignment Part 5 Modeling Credit Card Default Risk and Customer Profitability 2.txt

1 What is your predictive model? a. Describe the arithmetic clearly so that another learner could implement your model on new standardized input data if they wished. b. Give an example of the score you would assign the following applicant, whether they would be approved or reected for a credit card and why. !y predictive model is "#ncome $ %redit %ard Debt $ &uto Debt' ( "&ge)*ears at employer)*ears at address' the less outcome, the better. +hen, # did standardized Data by using "utput from my model per applicant $ mean'( -td.

and my conclusion is

When the output from standardized test / $0.0 that mean the ban2 can approve, while when the output from standardized test 3 $0.0 that mean the ban2 should reect.

 Give an example of the score you would assign the following applicant, whether they would be approved or reected for a credit card and why. Give an example of the score you would assign the following applicant, whether they would be approved or reected for a credit card and why. &ge4 $0.05 *ears at employer4 0.6 *ears at address4 $0.78 #ncome4 $0.68 %redit card debt4 0.19 &uto debt4 $0.05

&ccording to my model, the ban2 would approved this applicant which the test score is $0.:8, less than $0.. What would the the ban2� s average profit profit per applicant applicant be "net profits profits divided divided by 00' when using your predictive model on the +raining -et? +he ban2;s average profit is <,011 per applicant when using my predictive model on the training set. What is the incremental financial value per applicant of your model over no model on the +raining -et? +he incremental financial value per applicant of my model over no model on the +raining -et is <1 per applicant. What is the incremental financial value per applicant of your model over no model on the +raining -et? +he incremental financial value per applicant of my model over no model on the +raining -et is <1 per applicant.

=valuate your model on the +est -et data. >ow confident are you that your model does not over$fit the +raining -et data? &. %hoose between three broad degrees of confidence4 � very� � somewhat� or � not at all.� "ote that � not at all� is still an acceptable answer if you give persuasive reasons for why you chose this answer'. @. =xplain the evidence your degree of confidence is based upon. *our explanation should include the test set profits and training set profits per applicant. >ow much confidence to have in the model must relate to the relationship between the profits$per$applicant on the +raining -et and the +est -et +he relationship between the predictors and response is highly.