Abstract: We propose a new generic type of stochastic neurons, called $q$-neurons, that considers activation functions based on Jackson's $q$-derivatives, with stochastic parameters $q$. Our generalization of neural network architectures with $q$-neurons is shown to be both scalable and very easy to implement. We demonstrate experimentally consistently improved performances over state-of-the-art standard activation functions, both on training and testing loss functions.
Keywords: q-calculus, neural activation function
TL;DR: q-calculus helps build simple and scalable neural activation functions
Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/q-neurons-neuron-activations-based-on/code)
5 Replies
Loading