الملخص
Neural algorithmic reasoning, or aligning a neural network with an algorithmic paradigm, has emerged as an approach to solving polynomial-time-solvable and computationally harder combinatorial optimization problems. We propose a new message-passing framework based on the Chambolle-Pock Primal--Dual Hybrid Gradient (PDHG) method called \textsc{GraphPDHG} for solving general graph saddle-point problems. Theoretically, we show that \textsc{GraphPDHG} can efficiently solve a family of graph saddle-point problems by simulating PDHG. We also show that our network can learn an accelerated PDHG algorithm. Experimentally, we support our results on accelerated PDHG by evaluating the performance of our model as a learned warm start for second-order optimization techniques (SSNAL). We also show that alignment with PDHG leads to stronger size generalization than non-aligned graph neural network (GNN) baselines. Overall, we propose a novel architecture for solving a general family of optimization problems on graphs.
الكلمات المفتاحية
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Chen, S., He, J., Clougherty, C., Mishne, G., & Holtz, C. (2026). Neural Algorithmic Reasoning for Graph Saddle Point Problems. https://omanscience.com/ar/articles/neural-algorithmic-reasoning-for-graph-saddle-point-problems
MLA 9
Chen, Samantha, et al. "Neural Algorithmic Reasoning for Graph Saddle Point Problems." https://omanscience.com/ar/articles/neural-algorithmic-reasoning-for-graph-saddle-point-problems.
شيكاغو (المؤلف–التاريخ)
Chen, Samantha, Jesse He, Coleman Clougherty, Gal Mishne, and Chester Holtz. 2026. "Neural Algorithmic Reasoning for Graph Saddle Point Problems." https://omanscience.com/ar/articles/neural-algorithmic-reasoning-for-graph-saddle-point-problems.
هارفارد
Chen, S., He, J., Clougherty, C., Mishne, G. and Holtz, C. (2026) 'Neural Algorithmic Reasoning for Graph Saddle Point Problems', Available at: https://omanscience.com/ar/articles/neural-algorithmic-reasoning-for-graph-saddle-point-problems.
فانكوفر
Chen S, He J, Clougherty C, Mishne G, Holtz C. Neural Algorithmic Reasoning for Graph Saddle Point Problems. https://omanscience.com/ar/articles/neural-algorithmic-reasoning-for-graph-saddle-point-problems
IEEE
S. Chen, J. He, C. Clougherty, G. Mishne, and C. Holtz, "Neural Algorithmic Reasoning for Graph Saddle Point Problems," https://omanscience.com/ar/articles/neural-algorithmic-reasoning-for-graph-saddle-point-problems.