A brief history and future of credit scores
economist.com
A brief history and future of credit scores
1–10 of 72 posts
Re: A brief history and future of credit scores
#2Re: A brief history and future of credit scores
#3The key here being that non-financial data isn't actually useful in predicting ability to repay loans. I'm sure it'd be used somehow by modern financial institutions if it were predictive.
Re: A brief history and future of credit scores
#4http://archive.is/NtfoT
Thank you!
Re: A brief history and future of credit scores
#5The key here being that non-financial data isn't actually useful in predicting ability to repay loans. I'm sure it'd be used somehow by modern financial institutions if it were predictive.
I find this interesting because when I applied for a mortgage at Wells Fargo they asked me for my race in the application process.
Re: A brief history and future of credit scores
#6The key here being that non-financial data isn't actually useful in predicting ability to repay loans. I'm sure it'd be used somehow by modern financial institutions if it were predictive.
I find this interesting because when I applied for a mortgage at Wells Fargo they asked me for my race in the application process.
I get confused by those they claim some of these ML or NN algos are black boxes and so could be breaking the law. If you’re not inputting illegal info (like race, sex, national origin, religion, name, etc. or corollaries), then it’s not making its risk assessment on that basis. All you have to do is look at the inputs. It isn’t unknowable whether or not it’s breaking the law.
Re: A brief history and future of credit scores
#7The key here being that non-financial data isn't actually useful in predicting ability to repay loans. I'm sure it'd be used somehow by modern financial institutions if it were predictive.
I find this interesting because when I applied for a mortgage at Wells Fargo they asked me for my race in the application process.
They probably have to report on loan approval rates by race to demonstrate a lack of bias in their underwriting.
Re: A brief history and future of credit scores
#8Earlier quoted context omitted.
I find this interesting because when I applied for a mortgage at Wells Fargo they asked me for my race in the application process.
I get confused by those they claim some of these ML or NN algos are black boxes and so could be breaking the law. If you’re not inputting illegal info (like race, sex, national origin, religion, name, etc. or corollaries), then it’s not making its risk assessment on that basis. All you have to do is look at the inputs. It isn’t unknowable whether or not it’s breaking the law.
Read about disparate impact and you will see that it is possible to break the law without taking race, sex, national origin, etc as an input.
Re: A brief history and future of credit scores
#9Earlier quoted context omitted.
I find this interesting because when I applied for a mortgage at Wells Fargo they asked me for my race in the application process.
I get confused by those they claim some of these ML or NN algos are black boxes and so could be breaking the law. If you’re not inputting illegal info (like race, sex, national origin, religion, name, etc. or corollaries), then it’s not making its risk assessment on that basis. All you have to do is look at the inputs. It isn’t unknowable whether or not it’s breaking the law.
Let me try to clear your confusion: an input my seem innocent (e.g. zip code), but a zip code is likely to correlate to ethnicity and race in some regions. So even if the inputs seem legal, an ML model that’s sophisticated enough can derive illegal results that discriminate against certain populations.
Re: A brief history and future of credit scores
#10Earlier quoted context omitted.
I find this interesting because when I applied for a mortgage at Wells Fargo they asked me for my race in the application process.
I get confused by those they claim some of these ML or NN algos are black boxes and so could be breaking the law. If you’re not inputting illegal info (like race, sex, national origin, religion, name, etc. or corollaries), then it’s not making its risk assessment on that basis. All you have to do is look at the inputs. It isn’t unknowable whether or not it’s breaking the law.
Humans had plenty of algorithms to substantiate racial discrimination using nonobvious indicators. We can, inadvertantly or otherwise, easily program machines to do the same.