1 paper
Jonathan W. Siegel, Stephan Wojtowytsch
We consider gradient flow/gradient descent and heavy ball/accelerated gradient descent optimization for convex objective functions. In the gradient flow case, we prove the followin…