Use web service output in Machine Learning Studio

Question

I have a predictive experiment that has a webservice input and output. At the same time I need to execute an R script and the output of that script is a two columns table. (I am not sure where should I execute that script, in the same experiment or not? I'm new in Azure Machine Learning). Now I need to take the webservice output and look it up in the table resulted from R script. Is it even possible? The final result of the whole experiment should be the value taken from the table. Any advice will be helpful. Thank you in advance.

Answer 1

The Azure ML Studio (classic) webservice API is meant to be used to deploy "models". You give an input and a single prediction is returned. It sounds to me as if you'd like to take the output and also compare it to an intermediary table that's created in the pipeline.

I suggest:

1. Determine if the webservice can return a string that is the concatenated result of both outputs you desire, then parse the output into two objects downstream.
2. Find a way to make the webservice return exactly what you want instead of something that is one step away from what you want.
3. Look into Azure Machine Learning service, as ML Studio classic has not been under active development for some time now. It supports using R in the control plane and the data plane.