My project is binary classification prediction.I make standardise my numeric values and apply one-hot encoding to categorical values.I trained data and get 80% accuracy.I want to create web service and it has original data input and original data output but currently score model has standardised and one-hot encoded datas.So I concat two table which are original data and my prediction table's score probability,Scored label columns.Then I linked to web service output.I created web service but when I change feature datas Web service output always has same result.Where is my fault? At below,you can see my model.
Asked
Active
Viewed 42 times
-1
Sami Aydoğan
- 39
- 3
1 Answers
0
How are you verifying that the output is the same? By making an API call, right?
My understanding is that's the only way to use the web service output.
Anders Swanson
- 3,637
- 1
- 18
- 43
-
When you deploy web service in Azure Machine Learning Studio, you can test your new data with Request-Response tab and when you change your data, your prediction should change according to your input data.But when I change my input data,it doesn't change. – Sami Aydoğan Jun 10 '20 at 18:51