DeepLearning4.3B, Can we interpret the output as a probability?

views comments

We identify the conditions that enable us to interpret the output as a probability: Lot's of data; a flexible network; and the cross-entropy error function for classification tasks.