I'm working on a microsoft azure machine learning web service. I made a first version that takes a data set as an input and returns JSON that contains predictions. I want to add some python scripts that can process the predictions to change the output of the algorithm and replace it by another JSON file. I want also to add a script that treats other inputs that will not be used in the machine learning algorithm but will be treated in the output.
To be more clear : i have 5 attributes : x1, x2, x3, x4, x5 x1, x2 and x3 will be treated in the ML algorithm and return y : prediction the output that i want to have is some tips : if y meet some condition then output 1 (some string) but i want to process x4 x5 as well : if x4 and x5 meet some condition then output 2
the output will be : { output 1 : output 2 : } instead of { prediction : y }
I looked at the documentation of Azure but all i found is how to use python scripts to manipulate data frames. if some one has an idea on how to combine the ML Microsoft Azure Web Service and some python scripts to create a unique Cloud Based web service that would be great. Thank you
