none
How to input target variable for neural network regression model with multiple output nodes?

    Question

  • I am trying to create a pipeline that consists of a neural network that apply regression to the data, however I want to predict 4 values at the same time. In other words, the output layer consists of 4 nodes.

    My dataset that I use as input contains a target column containing lists, where each position of the list contains the target integer for one of the 4 nodes, like so:

         TargetValues
    0    [1, 2, 3, 3]
    1    [1, 2, 3, 4]
    2    [2, 3, 4, 6]
    3    [1, 3, 6, 7]

    I want to use the values in this column as my target values, however I'm not sure how to process that.

    The question:

    - Is it even possible to just do it like this?

    - If not, is there a way to transform the one column of lists to 4 columns of integers, like so:

         T1    T2    T3    T4
    0    1     2     3     3
    1    1     2     3     4
    2    2     3     4     6
    3    1     3     6     7

    - If not, is there another way to do this?

    Thanks in advance!



    • Edited by DevChuba Thursday, December 6, 2018 1:03 PM
    Thursday, December 6, 2018 12:52 PM

Answers

  • Hi,

    I think it should be convenient if you just use Python script/ R script to do the data format transfer and combine. Have you tried that?

    R : https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/execute-r-script

    Python: https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/execute-python-script

    Regards,

    Yutong

    Friday, December 7, 2018 6:53 AM
    Owner

All replies

  • Hi,

    I think it should be convenient if you just use Python script/ R script to do the data format transfer and combine. Have you tried that?

    R : https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/execute-r-script

    Python: https://docs.microsoft.com/en-us/azure/machine-learning/studio-module-reference/execute-python-script

    Regards,

    Yutong

    Friday, December 7, 2018 6:53 AM
    Owner
  • Ok, I will give that a try, thank you.
    Friday, December 7, 2018 8:04 AM