Δ ℚuantitative √ourney

On chain rule, computational graphs, and backpropagation

Here I'm going to revisit backpropagation theory by thinking about neural networks as computational graphs upon which we can easily visualize the chain rule to compute partial derivatives.