Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

tanh Neurons are Bayesian Decision Makers

: Bauckhage, Christian; Sifa, Rafet; Hecker, Dirk


Arai, Kohei:
Intelligent Systems and Applications. Proceedings of the 2020 Intelligent Systems Conference, IntelliSys 2020. Vol.3 : September 3-4, 2020, virtual conference
Cham: Springer Nature, 2021 (Advances in Intelligent Systems and Computing 1252)
ISBN: 978-3-030-55189-6 (Print)
ISBN: 978-3-030-55190-2 (Online)
ISBN: 978-3-030-55191-9
Intelligent Systems Conference (IntelliSys) <2020, Online>
Bundesministerium für Bildung und Forschung BMBF (Deutschland)
01-S18038A; ML2R
Conference Paper
Fraunhofer IAIS ()
hyperbolic tangent; decision boundaries; neural network

The hyperbolic tangent (tanh) is a traditional choice for the activation function of the neurons of an artificial neural network. Here, we go through a simple calculation that shows that this modeling choice is linked to Bayesian decision theory. Our brief, tutorial-like discussion is intended as a reference to an observation rarely mentioned in standard textbooks.