# What does the predict in Softmax mean?

I don’t understand what does the prediction mean in the result?
Prediction of what? Number 2 has the highest probability in the array 【1,3,4,2】？Why? It looks every number appears once each. Why 2 has higher possibility than others?
Newbie here. Anyone help explain. Many thanks!

1 Like

If you new to machine learning, you probably shouldn’t be taking this “deployment” specialization. It doesn’t teach the ML fundamentals.

This is just a math example of how the softmax() function works.

In application, we use softmax() when we have multiple classes, and the prediction for each example is the class with the highest value.

1 Like

Thank you for quick response. I do need to learn fundamentals.
Could you help explain why not 4 is the prediction? I see 4 is the highest value in the array.

1 Like

Ok. I got it. The output of the index number not the number itself.
Thanks. Problem solved.

1 Like

The prediction isn’t the highest value.
The prediction is the index of the location of the highest value.

The data set is constructed so that the indices (0 through n) represent the classes.

2 Likes