This project predicts mortgage loan approval result based on 2017 hawaii loan dataset obtained from Consumer Finance Protection Bureau
- Make your loan candidate following format: [loan_type, property_type, purpose, occupancy, amount, sex, income]
Numerical options:
- amount
- income
Categorical options
-------loan_type--------
Conventional
VA-guaranteed
FHA-insured
FSA/RHS-guaranteed
-------property_type--------
One-to-four family dwelling (other than manufactured housing)
Multifamily dwelling
Manufactured housing
-------purpose--------
Home purchase
Refinancing
Home improvement
-------occupancy--------
Owner-occupied as a principal dwelling
Not owner-occupied as a principal dwelling
Not applicable
-------sex--------
Male
Female
Information not provided by applicant in mail, Internet, or telephone application
Not applicable
Candiate example:
['Conventional', 'One-to-four family dwelling (other than manufactured housing)', 'Home improvement', 'Owner-occupied as a principal dwelling', 588, 'Male', 313],
- Put the candidate array into test.ipynb and hit run
df = pd.DataFrame(
[
['Conventional', 'One-to-four family dwelling (other than manufactured housing)', 'Home improvement', 'Owner-occupied as a principal dwelling', 588, 'Male', 313],
['Conventional', 'One-to-four family dwelling (other than manufactured housing)', 'Home improvement', 'Owner-occupied as a principal dwelling', 35, 'Male', 12]
], columns=['loan_type', 'property_type', 'purpose', 'occupancy', 'amount', 'sex', 'income'])
- Each candidate will return a 1 if approved or 0 if denied
# example return value from above
[1, 1] # this means both candidate were approved