Adaptive Weighting Strategies and Loss Balancing For Physics Informed Neural Network Solution of the Monge Ampere Partial Differential Equation

In this research we present a comprehensive computational framework for solving the nonlinear Monge-Ampère equation using an adaptive Physics-Informed Neural Network (PINN) approach. The Monge-Ampère equation is a fully nonlinear second-order partial differential equation which plays a fundamental role in optimal transport theory, differential geometry, image processing, and fluid mechanics. Despite its importance, numerical solutions remain challenging due to the equation's inherent nonlinearity, the requirement of convexity, and the existence of multiple solution branches. Traditional numerical methods, including finite differences and finite elements, often struggle with complex geometries, high-dimensional domains, and the enforcement of convexity constraints. We propose an adaptive PINN framework that addresses these limitations through a novel gradient magnitude balancing strategy. The neural network architecture employs a multilayer perceptron with tanh activation functions to approximate the solution on a square domain. Unlike supervised approaches that require labeled data, our unsupervised method incorporates the governing physics directly into the loss function through automatic differentiation, enabling mesh-free computation of the Hessian components. The total loss function comprises three components: a PDE residual loss enforcing the Monge-Ampère equation, a boundary loss enforcing Dirichlet conditions on the boundary, and a convexity loss penalizing negative eigenvalues of the Hessian matrix to ensure positive definiteness. The tanh activation produced an inverted (concave) solution because the convexity penalty wasn't strong enough to guide the optimizer to the correct branch.

Physics-Informed Neural Networks, Monge-Ampère Equation, Adaptive Loss Weighting, Gradient Magnitude Balancing, Convexity Enforcement

1. Yu, A., Wu, B., Zha, Z., Feng, X., & Sheen, D. (2026). Physics-Informed Neural Networks with Attention Feature Expansion for Monge-Ampère Equations. arXiv:2605.22115. [Google Scholar] [Crossref]

2. Review: Physics-Informed Neural Networks – Challenges in Loss Function Design and Geometric Integration. Mathematics, 13(20), 3289 (2025). [Google Scholar] [Crossref]

3. Lu, Y., Zhou, H., Yao, M., & Ren, Z. (2025). Enhanced Physics-Informed Neural Networks Using Adaptive Residual Weighting for Flow Simulations. GPPS TC-2025. [Google Scholar] [Crossref]

4. Wu, G., & Wu, Z. (2026). A Multi-Objective Optimization Framework for Adaptive Weighting in Physics-Informed Machine Learning. AAAI Conference on Artificial Intelligence, 40(32), 26885–26893. [Google Scholar] [Crossref]

5. Alex K., Yarin G., and Roberto C., (2018). Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. Proceedings of the IEEE conference on computer vision and pattern recognition (CVPR), pages 7482-7491. [Google Scholar] [Crossref]

6. Liu, L., & Wang, S. (2025). Adaptive Loss Balancing in PINNs. Review of Loss Function Design, Mathematics 13(20). [Google Scholar] [Crossref]

7. Wu, G., Vellayappan, K., Shi, Y., & Wu, Z. (2026). Pareto-Guided Regional Sampling for PINNs. ACS Industrial & Engineering Chemistry Research. [Google Scholar] [Crossref]

8. Pancheng N., Yongming C., Jun G., Yuqian Z., (2025). Improved Adaptive Weighting with Upper Bounds for PINNs. arXiv:2407.19421. [Google Scholar] [Crossref]

9. Hybrid Backpropagation Neural Network with Improved PSO for Monge-Ampère Equation. (2025). GPPS TC-2025. [Google Scholar] [Crossref]

10. Cao, K., Ding, X., Zhao, J., & Feng, X. (2025). Self-learning multi-head weight and enhanced physics-informed residual connection neural networks. Physics of Fluids, 37(4), 046121. https://doi.org/10.1063/5.0260860 [Google Scholar] [Crossref]

• Interplay of Students’ Emotional Intelligence and Attitude toward Mathematics on Performance in Grade 10 Algebra
• Numerical Simulation of Fitzhugh-Nagumo Dynamics Using a Finite Difference-Based Method of Lines
• Fixed Point Theorem in Controlled Metric Spaces
• Usage of Moving Average to Heart Rate, Blood Pressure and Blood Sugar
• Exploring Algebraic Topology and Homotopy Theory: Methods, Empirical Data, and Numerical Examples