Towards out of distribution generalization for problems in mechanics

Abstract

There has been a massive increase in research interest towards applying data driven methods to problems in mechanics. While traditional machine learning (ML) methods have enabled many breakthroughs, they rely on the assumption that the training (observed) data and testing (unseen) data are independent and identically distributed (i.i.d). Thus, traditional ML approaches often break down when applied to real world mechanics problems with unknown test environments and data distribution shifts. In contrast, out-of-distribution (OOD) generalization assumes that the test data may shift (i.e., violate the i.i.d. assumption). To date, multiple methods have been proposed to improve the OOD generalization of ML methods. However, because of the lack of benchmark datasets for OOD regression problems, the efficiency of these OOD methods on regression problems, which dominate the mechanics field, remains unknown. To address this, we investigate the performance of OOD generalization methods for regression problems in mechanics. Specifically, we identify three OOD problems: covariate shift, mechanism shift, and sampling bias. For each problem, we create two benchmark examples that extend the Mechanical MNIST dataset collection, and we investigate the performance of popular OOD generalization methods on these mechanics-specific regression problems. Our numerical experiments show that in most cases, while the OOD generalization algorithms perform better compared to traditional ML methods on these OOD problems, there is a compelling need to develop more robust OOD generalization methods that are effective across multiple OOD scenarios. Overall, we expect that this study, as well as the associated open access benchmark datasets, will enable further development of OOD generalization methods for mechanics specific regression problems.

References in corpus (19)

• Explaining and Harnessing Adversarial Examples
• Integrating Machine Learning and Multiscale Modeling: Perspectives, Challenges, and Opportunities in the Biological, Biomedical, and Behavioral Sciences
• Opportunities and Challenges for Machine Learning in Materials Science
• Generalized Out-of-Distribution Detection: A Survey
• WILDS: A Benchmark of in-the-Wild Distribution Shifts
• Towards Out-Of-Distribution Generalization: A Survey
• An Investigation of Why Overparameterization Exacerbates Spurious Correlations
• Forward and Inverse Design of Kirigami via Supervised Autoencoder
• Efficient Domain Generalization via Common-Specific Low-Rank Decomposition
• Understanding the Failure Modes of Out-of-Distribution Generalization
• An Ensemble of Simple Convolutional Neural Network Models for MNIST Digit Recognition
• Towards a Theoretical Framework of Out-of-Distribution Generalization
• A fourth-order phase-field fracture model: Formulation and numerical solution using a continuous/discontinuous Galerkin method
• Enhancing Mechanical Metamodels with a Generative Model-Based Augmented Training Dataset
• Kernelized Heterogeneous Risk Minimization
• On the Impact of Spurious Correlation for Out-of-distribution Detection
• Understanding and Testing Generalization of Deep Networks on Out-of-Distribution Data
• Bayesian OOD detection with aleatoric uncertainty and outlier exposure
• Predicting Mechanically Driven Full-Field Quantities of Interest with Deep Learning-Based Metamodels

Cited by in corpus (2)

• Physics-Informed Kolmogorov-Arnold Networks for multi-material elasticity problems in electronic packaging
• Uncertainty in AI-driven Monte Carlo simulations