Table 9 Performance comparison of the predictive solution and predictive parameter of Eq. (41).

From: Data-driven solutions and parameter estimations of a family of higher-order KdV equations based on physics informed neural networks

Case

PINN(\(\tanh\))

PINN(\(\sin\)e)

Loss

3.8809 × 10 −4

6.7632 × 10−5

\(\mathbb {L}_{2}\) Error u

9.5180 × 10−3

2.3708 × 10−3

Error \(\alpha\)

1.1878%

0.1151%

\(\alpha\)

1.0119

0.9988

Back to article page