On the application of Physically-Guided Neural Networks with Internal Variables to Continuum Problems, has been published in Mechanics of Materials.

Our members Rubén Muñoz Sierra, Jacobo Ayensa Jiménez and Manuel Doblaré have recently published a new research paper!

The work, called On the application of Physically-Guided Neural Networks with Internal Variables to Continuum Problems, has been published in Mechanics of Materials.

 

In this work they present a framework to solve continuum problems using machine learning that incorporates physical principles. This framework offers both predictive and explanatory capacity from available data. They have tested the method capacity to deal with heterogeneous, anisotropic and nonlinear materials.

 

In machine learning, the context of the models is crucial in order to create reliable and trustworthy models. Adding to them physical context as restrictions that are grounded in the real world ensures that the model gives realistic predictions. It also makes them less of a black box and more of a grey box, giving us the ability to explain how the model has made these predictions.

If you are interested and want to learn more about the job, you can find it here: https://lnkd.in/dJkWfdEt

 

Tissue Microenvironment lab - Universidad de Zaragoza | We are us and our circumstances (José Ortega y Gasset)

LEGAL

Privacy

 

CONTACT

tmelab@unizar.es

Rio Ebro Campus, Mariano Esquillor street w/o number, R&D Builiding, block 4
(50018 Zaragoza)

+34 876 55 54 66

2020 TME LAB © Designed by Cuidado con el Cactus