Why is automatic differentiation still used, if today's computers can calculate symbolic derivatives quite fast?
How can the sum of squared errors have negative gradient if it's defined as the squared of the error?
What does it mean by "zeros the networks parameters gradients" in the context of training a neural network?
Mathematically speaking, Is it only the product operation used in the chain rule causing the vanishing or exploding gradient?
In multilayer perceptron neural networks, are the names "delta", "gradient" and "error" all the same thing? or not?
Which is more popular/common way of representing a gradient in AI community: as a row or column vector?
Only top scored, non community-wiki answers of a minimum length are eligible