Robust Generalization with Adaptive Optimal Transport Priors for Decision-Focused Learning
Not on the map. Why?
Abstract (arXiv)
Few-shot learning requires models to generalize under limited supervision while remaining robust to distribution shifts. Existing Sinkhorn Distributionally Robust Optimization (DRO) methods provide theoretical guarantees but rely on a fixed reference distribution, which limits their adaptability. We propose a Prototype-Guided Distributionally Robust Optimization (PG-DRO) framework that learns class-adaptive priors from abundant base data via hierarchical optimal transport and embeds them into the Sinkhorn DRO formulation. This design enables few-shot information to be organically integrated into producing class-specific robust decisions that are both theoretically grounded and efficient, and further aligns the uncertainty set with transferable structural knowledge. Experiments show that PG-DRO achieves stronger robust generalization in few-shot scenarios, outperforming both standard learners and DRO baselines.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 42 of its 60 references that are in the map. First posted on arXiv 1 Feb 2026.
Builds on
42 of 60 references.
References · 42
- 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
- 2017Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networkscited by 4,861
- 2017Prototypical Networks for Few-shot Learningcited by 3,001
- 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
- 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,845
- 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,277
- 2016Matching Networks for One Shot Learningcited by 2,246
- 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,366
- 2018CosFace: Large Margin Cosine Loss for Deep Face Recognitioncited by 968
- 2018Computational Optimal Transportcited by 765
- 2015Optimal Transport for Domain Adaptationcited by 457
- 2018Meta-Learning for Semi-Supervised Few-Shot Classificationcited by 456
30 more
- 2018Adversarially Robust Generalization Requires More Datacited by 318
- 2015Data-driven Distributionally Robust Optimization Using the Wasserstein Metric: Performance Guarantees and Tractable Reformulationscited by 285
- 2024Reinforcement learningcited by 236
- 2013Fast Computation of Wasserstein Barycenterscited by 216
- 2018DeepJDOT: Deep Joint Distribution Optimal Transport for Unsupervised Domain Adaptationcited by 212
- 2019Generalized Sliced Wasserstein Distancescited by 205
- 2016Distributionally Robust Stochastic Optimization with Wasserstein Distancecited by 182
- 2016Variance-based regularization with convex objectivescited by 153
- 2018Gromov-Wasserstein Alignment of Word Embedding Spacescited by 119
- 2018Generative Modeling using the Sliced Wasserstein Distancecited by 108
- 2020Graph Optimal Transport for Cross-Domain Alignmentcited by 103
- 2020Adversarial Distributional Training for Robust Deep Learningcited by 79
- 2018Sliced-Wasserstein Flows: Nonparametric Generative Modeling via Optimal Transport and Diffusionscited by 77
- 2019Distributionally Robust Optimization and Generalization in Kernel Methodscited by 73
- 2018Optimal Transport for Multi-source Domain Adaptation under Target Shiftcited by 67
- 2022Unified Optimal Transport Framework for Universal Domain Adaptationcited by 57
- 2019Hierarchical Optimal Transport for Document Representationcited by 53
- 2020Domain-Adaptive Few-Shot Learningcited by 40
- 2021On Fast Adversarial Robustness Adaptation in Model-Agnostic Meta-Learningcited by 29
- 2019Confidence Regions in Wasserstein Distributionally Robust Estimationcited by 24
- 2021Learning Prototype-oriented Set Representations for Meta-Learningcited by 19
- 2022Distributionally Robust End-to-End Portfolio Constructioncited by 17
- 2024End-to-end Conditional Robust Optimizationcited by 17
- 2023Wasserstein distributional robustness of neural networkscited by 15
- 2022Adaptive Distribution Calibration for Few-Shot Learning with Hierarchical Optimal Transportcited by 14
- 2023Train Hard, Fight Easy: Robust Meta Reinforcement Learningcited by 11
- 2023Adversarial Robustness of Prompt-based Few-Shot Learning for Natural Language Understandingcited by 8
- 2024Differentiable Distributionally Robust Optimization Layerscited by 8
- 2019Adversarially Robust Few-Shot Learning: A Meta-Learning Approachcited by 6
- 2023Domain Adaptive Few-Shot Open-Set Learningcited by 4
Built on by
Citing papers per year: 2026: 2.
2 papers.