we have data from Prosper which was founded in 2005 as the first peer-to-peer lending marketplace in the United States. and the data of about 110,000 loans from 2005-2014 and the data contains 'Term', Loan Status, Borrower Rate, Estimated Loss, Estimated Return, loan orginal amount and much more.
you can download it from Here
'EstimatedLoss' and 'EstimatedReturn' dimensions are all highly correlated with one another, and all of them are also highly correlated with 'BorrowerRate' Since BorrowerRate is a measure of loan cost, it's not too surprising that the overall trend between BorrowerRate and the two individual estimations.
the best loan cost features (propserRatting AA, term 12, in the Q1 of the year, for customers(brorrwer) income +100,000 USD) seem to get the Lowest cost of the loan.
we can see that
Crate takes the most loans even when The terms, income range or Quarter of year vary.
the Estimatedloss increases as borrower_rate increases, as well as the lowesr borrower rate for the people withincome +100,000\$ and the Estimated Loss is none for those people.
For the presentation, I focus on the insights that helping person to choose When and how to get a loan with the least interest rate possible as :
What affects the interest rate.
the time best time to get alone.
best length of the loan expressed in months.
- python 3
- jupyter notebooks
- anaconda
- pandas
- NumPy
- Matplotlib
- seaborn