Enforcing convex constraints in Graph Neural Networks
Not on the map. Why?
Abstract (arXiv)
Many machine learning applications require outputs that satisfy complex, dynamic constraints. This task is particularly challenging in Graph Neural Network models due to the variable output sizes of graph-structured data. In this paper, we introduce ProjNet, a Graph Neural Network framework which satisfies input-dependant constraints. ProjNet combines a sparse vector clipping method with the Component-Averaged Dykstra (CAD) algorithm, an iterative scheme for solving the best-approximation problem. We establish a convergence result for CAD and develop a GPU-accelerated implementation capable of handling large-scale inputs efficiently. To enable end-to-end training, we introduce a surrogate gradient for CAD that is both computationally efficient and better suited for optimization than the exact gradient. We validate ProjNet on four classes of constrained optimisation problems: linear programming, two classes of non-convex quadratic programs, and radio transmit power optimization, demonstrating its effectiveness across diverse problem settings.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 30 of its 70 references that are in the map. First posted on arXiv 13 Oct 2025.
Builds on
30 of 70 references.
References · 30
- 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
- 2021Convex optimizationcited by 2,515
- 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,316
- 2013Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computationcited by 2,079
- 2021How Attentive are Graph Attention Networks?cited by 835
- 2016CVXPY: A Python-Embedded Modeling Language for Convex Optimizationcited by 641
- 2019Surrogate Gradient Learning in Spiking Neural Networkscited by 626
- 2017OptNet: Differentiable Optimization as a Layer in Neural Networkscited by 599
- 2019Differentiable Convex Optimization Layerscited by 465
- 2019Exact Combinatorial Optimization with Graph Convolutional Neural Networkscited by 318
- 2017OSQP: An Operator Splitting Solver for Quadratic Programscited by 161
- 2019Optimal Wireless Resource Allocation with Random Edge Graph Neural Networkscited by 90
18 more
- 2020Lagrangian Duality for Constrained Deep Learningcited by 74
- 2018Differential Properties of Sinkhorn Approximation for Learning with Wasserstein Distancecited by 56
- 2022Equivariant Descriptor Fields: SE(3)-Equivariant Energy-Based Models for End-to-End Visual Robotic Manipulation Learningcited by 56
- 2024Clarabel: An interior-point solver for conic programs with quadratic objectivescited by 44
- 2023A GNN-Guided Predict-and-Search Framework for Mixed-Integer Linear Programmingcited by 41
- 2022On Representing Linear Programs by Graph Neural Networkscited by 39
- 2021Practical Large-Scale Linear Programming using Primal-Dual Hybrid Gradientcited by 32
- 2023RAYEN: Imposition of Hard Convex Constraints on Neural Networkscited by 30
- 2018Learning Permutations with Sinkhorn Policy Gradientcited by 26
- 2024Learning Constrained Optimization with Deep Augmented Lagrangian Methodscited by 16
- 2024LinSATNet: The Positive Linear Satisfiability Neural Networkscited by 16
- 2015Gunrock: A High-Performance Graph Processing Library on the GPUcited by 14
- 2023A New Computationally Simple Approach for Implementing Neural Networks with Output Hard Constraintscited by 12
- 2024Elucidating the theoretical underpinnings of surrogate gradient learning in spiking neural networkscited by 11
- 2023PIQP: A Proximal Interior-Point Quadratic Programming Solvercited by 9
- 2020Sample-Specific Output Constraints for Neural Networkscited by 7
- 2025Towards graph neural networks for provably solving convex optimization problemscited by 4
- 2018Algorithms and Convergence Results of Projection Methods for Inconsistent Feasibility Problems: A Reviewcited by 1
Built on by
Citing papers per year: 2026: 2.
2 papers.