paper-with-me

Papers

Understanding Generalization in Physics Informed Models through Affine Variety Dimensions

2025-01-31 · Takeshi Koshizuka, Issei Sato

In recent years, physics-informed machine learning has gained significant attention for its ability to enhance statistical performance and sample efficiency by integrating physical structures into machine learning models. These structures, such as differential equations, conservation laws, and symmetries, serve as inductive biases that can improve the generalization capacity of the hybrid model. However, the mechanisms by which these physical structures enhance generalization capacity are not fully understood, limiting the ability to guarantee the performance of the models. In this study, we show that the generalization performance of linear regressors incorporating differential equation structures is determined by the dimension of the associated affine variety, rather than the number of parameters. This finding enables a unified analysis of various equations, including nonlinear ones. We introduce a method to approximate the dimension of the affine variety and provide experimental evidence to validate our theoretical insights.

📄 PDF Abstract BibTeX arXiv:2501.18879

Code (0)

등록된 구현이 없습니다.

Tasks

Physics-informed machine learning

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음

Similar Papers 제목 키워드 기반

What do physics-informed DeepONets learn? Understanding and improving training for scientific computing applications

2024-11-27 · Emily Williams, Amanda Howard, Brek Meuris, Panos Stinis

Physics-informed deep operator networks (DeepONets) have emerged as a promising approach toward numerically approximating the solution of partial differential equations (PDEs). In this work, we aim to develop further und…

Transfer Learning

Evolutionary Optimization of Physics-Informed Neural Networks: Evo-PINN Frontiers and Opportunities

2025-01-11 · Jian Cheng Wong, Abhishek Gupta, Chin Chun Ooi, Pao-Hsiung Chiu 외

Deep learning models trained on finite data lack a complete understanding of the physical world. On the other hand, physics-informed neural networks (PINNs) are infused with such knowledge through the incorporation of ma…

Evolutionary AlgorithmsModel OptimizationSurvey

Planning Safety Trajectories with Dual-Phase, Physics-Informed, and Transportation Knowledge-Driven Large Language Models

2025-04-06 · Rui Gan, Pei Li, Keke Long, Bocheng An 외

Foundation models have demonstrated strong reasoning and generalization capabilities in driving-related tasks, including scene understanding, planning, and control. However, they still face challenges in hallucinations, …

Computational EfficiencyGeneral KnowledgeLarge Language ModelScene Understanding+1

Physics-Informed Computer Vision: A Review and Perspectives

2023-05-29 · Chayan Banerjee, Kien Nguyen, Clinton Fookes, George Karniadakis

The incorporation of physical information in machine learning frameworks is opening and transforming many application domains. Here the learning process is augmented through the induction of fundamental knowledge and gov…

Inductive BiasPhysics-informed machine learningSystematic Literature Review

When Do Extended Physics-Informed Neural Networks (XPINNs) Improve Generalization?

2021-09-20 · Zheyuan Hu, Ameya D. Jagtap, George Em Karniadakis, Kenji Kawaguchi

Physics-informed neural networks (PINNs) have become a popular choice for solving high-dimensional partial differential equations (PDEs) due to their excellent approximation power and generalization ability. Recently, Ex…