Using Feed Forward Neural Network to Solve Eigenvalue Problems

Table 2

Accuracy of solutions for Example 1.

The error 𝐸(𝑥)=|𝑦𝑡(𝑥)−𝑦𝑎(𝑥)|, where 𝑦𝑡(𝑥) is computed by the following training algorithms
Trainlm	Trainbfg	Trainbr
2.220446049250313𝑒−16	3.471978260449760𝑒−10	2.708944180085382𝑒−14
3.330669073875470𝑒−16	3.273391557812033𝑒−10	3.868017017794045𝑒−13
1.314053033052431𝑒−16	4.320360939662205𝑒−10	8.317256171980203𝑒−12
1.286829394127678𝑒−16	2.001726400755920𝑒−05	2.063986935888358𝑒−10
1.424081636936567𝑒−07	9.133774782171392𝑒−08	2.825134249813369𝑒−09
1.702890548670126𝑒−16	4.773932412291132𝑒−09	1.930348375213914𝑒−08
4.440892063174340𝑒−16	1.076577582637128𝑒−08	6.462875423718373𝑒−08
2.716475933084439𝑒−10	6.335612035002214𝑒−09	9.347123508529620𝑒−08
1.559072870804812𝑒−10	2.091692152816904𝑒−08	2.220824357745954𝑒−08
1.710883542650648𝑒−53	6.814833142243515𝑒−10	6.371296168428131𝑒−08
1.589879339292111𝑒−10	8.929697603576870𝑒−08	4.040193102294865𝑒−08