2
$\begingroup$

Newton Method is an optimization equation to find minimum point better than Gradient descent. Now we use momentum with gradient for smoothly. Why don't we use the Newton method?

$\endgroup$

1 Answer 1

2
$\begingroup$

Newton's method involves the calculation of the inverse of Hessian matrix, which is far too expensive to calculate in many cases in deep learning. That is why other approximation methods are preferred.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.