M-to-N Regression Prediction Demo Code

Hey

I am trying to train a SDCA regression (like the Taxi-Fair example) but instead of predicting a single output value it should predict a vector of 5 values (I have a vector of 128 input floats and a vector of 5 output floats that I try to learn).

I can't figure out how to get this to work and I can not find any online examples that do that either. Can you point me to some dummy code on how to get this to work? I think my error is with how I setup my data processing pipeline.

Sorry, I can't share any code here but if you can just point me to some example code on how to achieve that would be great.

Oli