English
Related papers

Related papers: Overcoming the Loss Conditioning Bottleneck in Opt…

200 papers

While Physics-Informed Neural Networks (PINNs) are powerful for solving Partial Differential Equations (PDEs), their training is often paralyzed by gradient pathology. The gradients from the PDE residuals and boundary constraints oppose…

Machine Learning · Computer Science 2026-05-26 Yichen Luo , Peiyu Zhu , Dongxiao Hu , Jia Wang , Tailin Wu , Dapeng Lan , Yu Liu , Zhibo Pang

In many scientific and engineering (e.g., physical, biochemical, medical) practices, data generated through expensive experiments or large-scale simulations, are often sparse and noisy. Physics-informed neural network (PINN) incorporates…

Machine Learning · Computer Science 2025-10-21 Kai-liang Lu

Partial differential equations (PDEs) provide a mathematical foundation for simulating and understanding intricate behaviors in both physical sciences and engineering. With the growing capabilities of deep learning, data$-$driven approaches…

Machine Learning · Computer Science 2025-10-14 Narayan S Iyer , Bivas Bhaumik , Ram S Iyer , Satyasaran Changdar

(Partial) differential equations (PDEs) are fundamental tools for describing natural phenomena, making their solution crucial in science and engineering. While traditional methods, such as the finite element method, provide reliable…

Machine Learning · Computer Science 2025-03-11 Viggo Moro , Luiz F. O. Chamon

We propose a meta-learning technique for offline discovery of physics-informed neural network (PINN) loss functions. We extend earlier works on meta-learning, and develop a gradient-based meta-learning algorithm for addressing diverse task…

Machine Learning · Computer Science 2022-04-13 Apostolos F Psaros , Kenji Kawaguchi , George Em Karniadakis

Classical stochastic gradient methods are well suited for minimizing expected-value objective functions. However, they do not apply to the minimization of a nonlinear function involving expected values or a composition of two expected-value…

Machine Learning · Statistics 2014-11-17 Mengdi Wang , Ethan X. Fang , Han Liu

Physics-Informed Neural Networks (PINN) are algorithms from deep learning leveraging physical laws by including partial differential equations together with a respective set of boundary and initial conditions as penalty terms into their…

Machine Learning · Computer Science 2025-09-30 Rafael Bischof , Michael Kraus

Physics-informed neural networks (PINNs) are a promising approach that combines the power of neural networks with the interpretability of physical modeling. PINNs have shown good practical performance in solving partial differential…

Statistics Theory · Mathematics 2026-01-26 Nathan Doumèche , Gérard Biau , Claire Boyer

Despite an extensive body of literature on deep learning optimization, our current understanding of what makes an optimization algorithm effective is fragmented. In particular, we do not understand well whether enhanced optimization…

Machine Learning · Computer Science 2024-03-04 Toki Tahmid Inan , Mingrui Liu , Amarda Shehu

Over the past few years, neural network methods have evolved in various directions for approximating partial differential equations (PDEs). A promising new development is the integration of neural networks with classical numerical…

Numerical Analysis · Mathematics 2025-07-10 Georgios Grekas , Charalambos G. Makridakis , Tristan Pryer

Physics Informed Neural Networks (PINNs) often exhibit failure modes in which the PDE residual loss converges while the solution error stays large, a phenomenon traditionally blamed on local optima separated from the true solution by steep…

Machine Learning · Computer Science 2025-12-01 Chenhui Xu , Dancheng Liu , Amir Nassereldine , Jinjun Xiong

Physics-Informed Neural Networks (PINNs) have revolutionized the computation of PDE solutions by integrating partial differential equations (PDEs) into the neural network's training process as soft constraints, becoming an important…

Machine Learning · Computer Science 2026-02-10 Elham Kiyani , Khemraj Shukla , Jorge F. Urbán , Jérôme Darbon , George Em Karniadakis

Large-scale constrained optimization problems are at the core of many tasks in control, signal processing, and machine learning. Notably, problems with functional constraints arise when, beyond a performance{\nobreakdash-}centric goal…

Optimization and Control · Mathematics 2025-05-15 Antesh Upadhyay , Sang Bin Moon , Abolfazl Hashemi

Physics-Informed Neural Networks (PINNs) provide a powerful and general framework for solving Partial Differential Equations (PDEs) by embedding physical laws into loss functions. However, training PINNs is notoriously difficult due to the…

Machine Learning · Computer Science 2025-10-09 Kang An , Chenhao Si , Ming Yan , Shiqian Ma

Neural differential equations offer a powerful framework for modeling continuous-time dynamics, but forecasting stiff biophysical systems remains unreliable. Standard Neural ODEs and physics informed variants often require orders of…

Machine Learning · Computer Science 2025-11-18 Kamalpreet Singh Kainth , Prathamesh Dinesh Joshi , Raj Abhijit Dandekar , Rajat Dandekar , Sreedat Panat

This work considers stochastic Galerkin approximations of linear elliptic partial differential equations (PDEs) with stochastic forcing terms and stochastic diffusion coefficients, that cannot be bounded uniformly away from zero and…

Numerical Analysis · Mathematics 2026-01-12 Fabio Musco , Andrea Barth

We introduce Multi-Iteration Stochastic Optimizers, a novel class of first-order stochastic methods that control the relative $L^2$ error using successive control variates along the iteration path. By exploiting correlations between…

Optimization and Control · Mathematics 2026-05-28 Andre Carlon , Luis Espath , Rafael Lopez , Raul Tempone

The stochastic gradient descent (SGD) method is a widely used approach for solving stochastic optimization problems, but its convergence is typically slow. Existing variance reduction techniques, such as SAGA, improve convergence by…

Optimization and Control · Mathematics 2025-11-21 Fabio Nobile , Matteo Raviola , Nathan Schaeffer

We leverage Physics-Informed Neural Networks (PINNs) to learn solution functions of parametric Navier-Stokes Equations (NSE). Our proposed approach results in a feasible optimization problem setup that bypasses PINNs' limitations in…

Computational Engineering, Finance, and Science · Computer Science 2024-02-06 M. Naderibeni , M. J. T. Reinders , L. Wu , D. M. J. Tax

We discover restrained numerical instabilities in current training practices of deep networks with stochastic gradient descent (SGD), and its variants. We show numerical error (on the order of the smallest floating point bit and thus the…

Machine Learning · Computer Science 2024-06-13 Yuxin Sun , Dong Lao , Ganesh Sundaramoorthi , Anthony Yezzi