Improving Underwriting of Branch Users
Mitchell Reynolds
May 3, 2018
Project objective:
Provide Insights that guide the underwriting process
Summary of Insights
Time Series
A key determining factor for repayment is when the loan is disbursed.
High Level Features
Of all easily derived variables, it seems that how “popular” a person is predicts if they will repay or default.
Natural Language Processing
Two different approaches
Several words seem to shed light onto defaults and repayments.
Time Series
All loans defaulted if they were disbursed before May 25th, then defaults greatly reduce.
Possible Explanation: Repayment time is within a week for Branch and these users get paid around the first of the month.
High Level Features - Logit
Most Important (Pearson R)
After date of disbursement, the highest Pearson R score is the unique count of each variable below for each user:
SMS Derived Variables:
MULTICOLLINEARITY
Natural Language Processing (NLP) Crash Course
NLP Approach 1 - Combined Messages
NLP Approach 1 - Combined Messages
Word Counts
TFIDF
Combined Messages
Milk, number, want, and tissue seem to move with defaults.
No English words made it into the top 10 for repayment
NLP Approach 2 - Each Sentence
NLP Approach 2 - Each Sentence
Word Counts
TFIDF
Each Sentence
This doesn’t make a lot of sense to me…
I think the models might be overfitting to certain words.
Model Comparison
High Level Logit
NLP Logit
Stacked Logit
Thank You!
Questions?
Appendix
Time Series
Most phone calls happen in May then decay thereafter.