Skip to main content
. 2023 Jul 24;14:4440. doi: 10.1038/s41467-023-40192-2

Fig. 6. Training a CNN to predict continuous values of area momentum.

Fig. 6

Top panel: training and validation loss as a function of epoch number. Bottom panel: mean and standard deviation (circular markers and error bars, respectively) of the CNN predictions on the test set. The number of distinct values of momentum to predict is equal to the number of dots in the grid of Fig. 2a and adds up to 36.