Create a model to estimate the probability of use the credit card, using max 50 variables.
It must be done using: Random Forest, Logistic Regression.
Sample database "[login to view URL]" with 7700 records and 397 variables.
The basis train can be: train 80% and test 20%
Target variable: FLG_ACTIVACION_0_12 (1 = used the card, 0 = not used).
- Just Jupiter notebook file with comments and commands.
PD. I attach an example for xgBoost (jupiter notebook as example).
5 freelancers are bidding on average $18 for this job
I'm well versed with machine learning, in a previous project I completed a similar project with XGBoost Feel free to ping me if my profile interests you.