HuHaigen / Adaptively-Customizing-Activation-FunctionsView on GitHub
To enhance the nonlinearity of neural networks and increase their mapping abilities between the inputs and response variables, activation functions play a crucial role to model more complex relationships and patterns in the data. In this work, a novel methodology is proposed to adaptively customize activation functions only by adding very few pa…
18Feb 20, 2024Updated 2 years ago

Alternatives and similar repositories for Adaptively-Customizing-Activation-Functions

Users that are interested in Adaptively-Customizing-Activation-Functions are comparing it to the libraries listed below

Sorting:

Are these results useful?