(There's only one at the moment; I'm just starting out!)
This zine covers mathematical derivations for the following:
- Binary classification using a neuron
- Non-linear classification with a Multilayer Perceptron
- Effect of adding layers to a Multilayer Perceptron.
It's intended to be a compilation of summary notes on these topics, especially as reference material for backpropagation in fully connected neural networks. Companion posts — which include code and demos — can be found here.
Link to zine here.