> Technically combining stochasticity with gradient descent is not pure gradient descent.
Sounds like you're picking a nit. Stochastic gradient descent is included in "gradient descent techniques."
Sounds like you're picking a nit. Stochastic gradient descent is included in "gradient descent techniques."
As for the domain knowledge - this is not specific to gradient descent either, genetic algorithms and their problem encoding can be adapted to a domain to minimise the search space considered - therefore improving performance of the algorithm.