Glossary term
Glossary term
Foundations
The "final" layer of a neural network. The output layer contains the prediction.
The following illustration shows a small deep neural network with an input layer, two hidden layers, and an output layer:
Created for this library
A churn modeling team uses a sigmoid output layer because the production system needs calibrated probabilities for prioritization.
A multi-class NLP team uses a softmax output layer over its 20 ticket categories.
A regression team uses a linear output layer for its sales forecast to allow unbounded continuous predictions.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License