Page 8 - summer of ai

8
DATA & ANALYTICS SUMMER OF AI JULY 2019
NeuroDecision®
When making a decision impacting a consumer’s financial well-being, it’s critical to explain and justify
the reasons underpinning the decisions.
Explainable AI and insights generation
Learning from data
Illuminating the black box means giving customers the key factors that underpin credit decisions and ethically expanding credit access for consumers
interest rate.
Matt Turner, a principal mathematical statistician who holds a PhD in mathematics and master’s degree in statistics, discusses the need for explainable AI, de-mystifies
other artificial intelligence techniques and shares how NeuroDecision Technology helps customers approve better loans for their clients whether a new homebuyer, small business owner or new mobile subscriber. Artificial Intelligence that learns from data like the explainable AI in NeuroDecision Technology will help customers improve outcomes for
their clients – especially those of us who are
at or near the edges – without impacting a customer’s risk profile.
According to Matt, the credit scoring models – like the ones that decide mortgage loans – may have been powered by a neural network: a machine learning technique that mimics the way the human brain processes information.
Matt details the way a neural network mimics the human brain. “Neural networks are a series of activation units called neurons that combine
The Equifax Data Science Lab invented NeuroDecision® Technology – the world’s first explainable AI patent in the credit industry – in order to help businesses, approve more customers while keeping risk levels constant, or decrease its delinquency rate while holding its approval rate constant.
According to the National Association of Realtors, if you purchased a home in 2018 in the United States, you were one of 6 million people who purchased a new or existing home. Summarizing the Quarterly Report on Household Debt and Credit released from the
Equifax Consumer Credit Panel and Federal Reserve Bank of New York, analysts found you likely had a credit score above 760 because nearly three out five borrowers approved for a mortgage then had excellent credit. The share of all mortgages given to people with excellent credit skyrocketed from 28% between 2001 and 2008 to 57% by the end of 2018.
However, do not think mortgage lenders
are only approving mortgages with people with excellent credit. By the end of 2018, banks originated $38.3 billion in mortgages for people with a credit score less than 660. Though these home buyers received a home mortgage, these consumers likely accepted a mortgage with more costly terms like a higher