DSN AI Bootcamp Qualification Hackathon by Data Science Nigeria
Predict customers who will default on a loan
888 data scientists enrolled, 522 on the leaderboard
Financial ServicesPredictionStructured
9 September—3 October
Ends in 8 days
Undefined Variables 48_49
published 12 Sep 2020, 21:37

Still thinking on what to do to these undefined variables. Any suggestion?

Why are they undefined?

Drop it and see the direction of yhur leaderboard score and then decide what to do with it.

Yes i agree! Its called backward/forward selection because machines tends to develop some patterns that seems to work out well but i have not proved that yet... so such approach is valid; But thats not the best way to work so far; its time consuming and in the place of deploying into production no room for such approach..


Kindly check there values if the same, if yes, drop one and if correlation seems to look alike drop that with lower correlation

if not then take the sum of the features to reduce dimensionality

Ok Agege 👌