Previous | Next --- Slide 12 of 27
Back to Lecture Thumbnails
seven

Move at the least the distance

ohmygearbox

sounds kind of like gradient descent

coolbreeze

First a larger step. Go to the estimated point where the distance is zero. Then take little steps

doberdog91

Fancier gradient descent algorithms like ADAM

Or if the distance function is analytic, compute its zero

alexz2

that sounds like gradient descent