I have to provide 2 models(1 has to neural net). It is for a card credit card fraud detection proj.

Andr3w_j_AAndr3w_j_A Member Posts: 3 Newbie
edited May 15 in Help

Here is the process i have come up with


Linear regression model


Neural Net
Struggling so bad so can anyone please help asap, please. It cannot even run as it is saying: "non-nominal label". Performance (classification) cannot handle numerical label. Can any one please help

Best Answer

  • BalazsBaranyBalazsBarany Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert Posts: 803 Unicorn
    Solution Accepted
    Hi!

    I don't see your label in the screenshot. What kind of attribute is it? The modeling algorithms and also the Performance operator need to match the label's type. 

    You should do Set Role before the cross validation, not inside it. The Performance operator needs to have both the prediction and the label attributes, but if you do it only in Training, it won't be applied in Testing. 

    Regards,
    Balázs
    Andr3w_j_A

Answers

Sign In or Register to comment.