CODEWITHSUNDEEP
python-3.xmachine-learningneat
Igor Markovic
Igor Markovic

Reputation: 185

NEAT multiple activation functions

I was wondering if a NEAT network could have multiple activation functions.

https://neat-python.readthedocs.io/en/latest/activation.html

For example two neurons as an output: one sigmoid and one selu.

I also think if I change it in the config file to selu for example, all neurons would have the selu activation function. Please correct me if I'm wrong.

I would like to set all activation functions to sigmoid for all layers and outputs and only one output to selu for example. Is this possible or am I just talking non-sense here?

Thanks in advance.

Upvotes: 2

Views: 491

Answers (0)

Related Questions