Error predicting attribute with missing values

pblack476pblack476 Member Posts: 83 Maven
edited December 2019 in Help
I have a time series and I want to predict a binary label I've created. But on AutoModel it throws out an error for every learner when I try to predict with a missing label. Is this usual behaviour or there might be something wrong here?

I mean... what is the point of predicting if you cannot predict on an unknown label value.

Best Answer

  • Options
    pblack476pblack476 Member Posts: 83 Maven
    Solution Accepted
    I figured out that I just had to train it on the know label data and apply it later on the unknown one. With Regression tasks I was used to just training with the unknown labels. Livin and learning
Sign In or Register to comment.