Fig. 7
From: Accelerated ensemble optimization using momentum methods
Final NPV (represented as dots) of each method (upper panel) and number of simulations (lower panel) from Fig. 6. Note that for Momentum and Nesterov, setting \(\beta =0\) yields the standard gradient descent method. Because the runs are run with five random seeds, the sample point is too low to display the result as a box-plot (as is done in Fig. 3).
