### References & Citations

# Computer Science > Computational Engineering, Finance, and Science

# Title: Neural Networks to solve Partial Differential Equations: a Comparison with Finite Elements

(Submitted on 10 Jan 2022 (v1), last revised 8 Apr 2022 (this version, v3))

Abstract: We compare the Finite Element Method (FEM) simulation of a standard Partial Differential Equation thermal problem of a plate with a hole with a Neural Network (NN) simulation. The largest deviation from the true solution obtained from FEM ($0.015$ for a solution on the order of unity) is easily achieved with NN too without much tuning of the hyperparameters. Accuracies below $0.01$ instead require refinement with an alternative optimizer to reach a similar performance with NN. A rough comparison between the Floating Point Operations values, as a machine-independent quantification of the computational performance, suggests a significant difference between FEM and NN in favour of the former. This also strongly holds for computation time: for an accuracy on the order of $10^{-5}$, FEM and NN require $54$ and $1100$ seconds, respectively. A detailed analysis of the effect of varying different hyperparameters shows that accuracy and computational time only weakly depend on the major part of them. Accuracies below $0.01$ cannot be achieved with the "adam'' optimizers and it looks as though accuracies below $10^{-5}$ cannot be achieved at all. In conclusion, the present work shows that for the concrete case of solving a steady-state 2D heat equation, the performance of a FEM algorithm is significantly better than the solution via networks.

## Submission history

From: Andrea Sacchetti [view email]**[v1]**Mon, 10 Jan 2022 10:34:57 GMT (2876kb,D)

**[v2]**Tue, 11 Jan 2022 07:12:38 GMT (2876kb,D)

**[v3]**Fri, 8 Apr 2022 05:32:02 GMT (1601kb)

Link back to: arXiv, form interface, contact.