Can I read Derivatives of Activation Functions on EtoBox?
Derivatives of Activation Functions by saheb_ju is a document available to read on EtoBox.
What is Derivatives of Activation Functions about?
The document describes code snippets for building and training a neural network model. It includes functions for: 1. Computing the derivatives of activation functions (diff_actFun) for sigmoid, tanh, and ReLU activations. 2. Implementing feedforward propagation through a 3-layer network to compute class probabilities. 3. Calculating the cross-entropy loss over predictions and true labels, plus optional regularization.
- Author
- saheb_ju
- Language
- EN