Why Azure ML , scored probability response greater than 1?

Question

I have setup an Azure ML experiment,using "Boosted Decision Tree Regression".

My evaluation results are as follows:

"Scored Labels": "N",
"Scored Probabilities": "0.023*************"

While testing the deployed web service with data, I am sometimes seeing a scored probability that is greater than 1!

"Scored Labels": "N", 
"Scored Probabilities": "1.144*************"

As per my understanding, probability of anything can never be grater than 1. What does a scored probability > 1 in this case mean?

Answer 1

It's related to Score Model and up to the type of your used model Boosted Decision Tree Regression .

Please refer to the subsection Results of the offical document Score Model , as below.

The score, or predicted value, can be in many different formats, depending on the model and your input data:

• For classification models, Score Model outputs a predicted value for the class, as well as the probability of the predicted value.
• For regression models , Score Model generates just the predicted numeric value .
• For image classification models, the score might be the class of object in the image, or a Boolean indicating whether a particular feature was found.

So in your case, the Scored Probabilities value is just the predicted numeric value, not the probability value.

If you want to get a value between 0 and 1, you can use the Normalize Data Module to change it to a 0-1 scale.