Gaussian (normal distribution) activation function

[damiano1996]

Describe the feature and the current behavior/state.
After this article I tried to implement a perceptron with an apical activation function similar to the real.
An approximation function of the apical was the gaussian, thus I implemented it.
In my fork you can find it.

Now, also the perceptron can learn the XOR logic function:

Relevant information

• Are you willing to contribute it? (yes/no): yes.
• Are you willing to maintain it going forward? (yes/no): no.
• Is there a relevant academic paper? (if so, where): I don't know.
• Is there already an implementation in another framework? (if so, where): I did not find anything.
• Was it part of tf.contrib? (if so, where): no.

Which API type would this fall under (layer, metric, optimizer, etc.)
activations.

Who will benefit with this feature?
researchers, developers.

Any other info.

[damiano1996]

What can I do? The code is just ready in my fork.

[seanpmorgan]

@damiano9669 Apologies for the delay in response. So I think this is an interesting idea, but I do not have a lot of intuition on how useful this is to the community at large. Without an academic paper describing the results and uses it's difficult to asses.

I would propose we leave this issue open and if community interests builds we could accept a PR for this. Alternatively if you proposed (and then included) a nice tutorial that shows its use and how it accomplishes something improved or novel then it would be a potential addition IMO.

[seanpmorgan]

TensorFlow Addons is transitioning to a minimal maintenance and release mode. New features will not be added to this repository. For more information, please see our public messaging on this decision:
TensorFlow Addons Wind Down

Please consider sending feature requests / contributions to other repositories in the TF community with a similar charters to TFA:
Keras
Keras-CV
Keras-NLP