I am learning to use machine learning with Mathematica. I have a problem that consists of 3xN input values that result in 2xN output values (N number of samples); however, the predict function seems to give only one output value. In this regard, it is possible to predict the outputs independently (i.e., Predict[{1 2 3}]=10-- three inputs and one output each time), but it is different from what the problem requires.
Therefore, I wonder if there is a way to get a self-consist result. To be more precise, I would like to consistently solve a problem like the one exemplified below.
So, I have 3xN input samples-> 2xN outputs (assuming N=5)
- 1 2 3 -> 10 20
- 2 3 1 -> 20 20
- 1 2 3 -> 20 30
- 4 5 6 -> 10 20
- 1 2 1 -> 20 20
In a way I can type:
to obtain:
Thank you for your help.