Importance of hyper-parameter optimization during training of physics-informed deep learning networks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lenau, Ashley, Dimiduk, Dennis M., Niezgoda, Stephen R.
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913358258634752
author Lenau, Ashley
Dimiduk, Dennis M.
Niezgoda, Stephen R.
author_facet Lenau, Ashley
Dimiduk, Dennis M.
Niezgoda, Stephen R.
contents Incorporating scientific knowledge into deep learning (DL) models for materials-based simulations can constrain the network's predictions to be within the boundaries of the material system. Altering loss functions or adding physics-based regularization (PBR) terms to reflect material properties informs a network about the physical constraints the simulation should obey. The training and tuning process of a DL network greatly affects the quality of the model, but how this process differs when using physics-based loss functions or regularization terms is not commonly discussed. In this manuscript, several PBR methods are implemented to enforce stress equilibrium on a network predicting the stress fields of a high elastic contrast composite. Models with PBR enforced the equilibrium constraint more accurately than a model without PBR, and the stress equilibrium converged more quickly. More importantly, it was observed that independently fine-tuning each implementation resulted in more accurate models. More specifically, each loss formulation and dataset required different learning rates and loss weights for the best performance. This result has important implications on assessing the relative effectiveness of different DL models and highlights important considerations when making a comparison between DL methods.
format Preprint
id arxiv_https___arxiv_org_abs_2405_08580
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Importance of hyper-parameter optimization during training of physics-informed deep learning networks
Lenau, Ashley
Dimiduk, Dennis M.
Niezgoda, Stephen R.
Materials Science
Data Analysis, Statistics and Probability
Incorporating scientific knowledge into deep learning (DL) models for materials-based simulations can constrain the network's predictions to be within the boundaries of the material system. Altering loss functions or adding physics-based regularization (PBR) terms to reflect material properties informs a network about the physical constraints the simulation should obey. The training and tuning process of a DL network greatly affects the quality of the model, but how this process differs when using physics-based loss functions or regularization terms is not commonly discussed. In this manuscript, several PBR methods are implemented to enforce stress equilibrium on a network predicting the stress fields of a high elastic contrast composite. Models with PBR enforced the equilibrium constraint more accurately than a model without PBR, and the stress equilibrium converged more quickly. More importantly, it was observed that independently fine-tuning each implementation resulted in more accurate models. More specifically, each loss formulation and dataset required different learning rates and loss weights for the best performance. This result has important implications on assessing the relative effectiveness of different DL models and highlights important considerations when making a comparison between DL methods.
title Importance of hyper-parameter optimization during training of physics-informed deep learning networks
topic Materials Science
Data Analysis, Statistics and Probability
url https://arxiv.org/abs/2405.08580