Comment by fordo on Everyday Lessons from High-Dimensional Optimization · 2020-06-14T15:30:45.074Z · LW · GW

Do you count gradient descent as a blackbox optimization, and isn't backpropagation guided by gradient (at least in ANN)?

Comment by fordo on Everyday Lessons from High-Dimensional Optimization · 2020-06-14T15:05:26.194Z · LW · GW

Care to elaborate how did you get to the formula O(1/sqrt(n))?