feed0 / german-credit

Credit Risk Modeling with Python
MIT License
0 stars 0 forks source link

One-hot-encode categorical features in logreg #5

Closed feed0 closed 2 weeks ago

feed0 commented 2 weeks ago

Before applying OHE it is necessary to preprocess some features as ordinal and adjust the classes labels such as in

BEFORE
1:Q - Status of existing checking account

- A11 :      ... <    0 DM
- A12 : 0 <= ... <  200 DM
- A13 :      ... >= 200 DM / salary assignments for at least 1 year

- A14 : no checking account
AFTER
1:O - Status of existing checking account

- A10 : no checking account

- A11 :      ... <    0 DM
- A12 : 0 <= ... <  200 DM
- A13 :      ... >= 200 DM / salary assignments for at least 1 year

Thus this issue was created: Ordinal features