New article: A note on data-driven methods for mechanical problems with non-unique solutions
New article: Romero, I. and Ortiz, M. (2026). A note on data-driven methods for mechanical problems with non-unique solutions, Meccanica, 61. (link).
In this article, we show that most of the usual machine learning models have a serious drawback when employed in nonlinear mechanics: they can not predict the bifurcation of solutions and therefore might incur in serious misrepresentations, even for the simplest problem.