Research map / Learning theory and algorithms / Prediction and decision losses
Distributionally Robust Graphical Models
Rizal Fathony, Ashkan Rezaei, Mohammad Ali Bashiri, Xinhua Zhang, Brian Ziebart
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Train with AGM's loss-aligned adversarial objective instead of CRF training | Structured prediction · Facial emotion intensity prediction, absolute unweighted loss | loss: 0.34 -> 0.33 | Improved |
| Train with AGM's loss-aligned adversarial objective instead of CRF training | Structured prediction · Facial emotion intensity prediction, weighted zero-one loss | loss: 0.32 -> 0.28 | Improved |
| Train with AGM's loss-aligned adversarial objective instead of CRF training | Structured prediction · Facial emotion intensity prediction, weighted absolute loss | loss: 0.36 -> 0.29 | Improved |
| Train with AGM's loss-aligned adversarial objective instead of CRF training | Structured prediction · Facial emotion intensity prediction, unweighted zero-one loss | loss: 0.32 -> 0.34 | Hurt |
| Train with AGM's loss-aligned adversarial objective instead of CRF training | Structured prediction · Facial emotion intensity prediction, squared unweighted loss | loss: 0.38 -> 0.38 | No clear effect |
| Train with AGM's loss-aligned adversarial objective instead of CRF training | Structured prediction · Facial emotion intensity prediction, average loss | loss: 0.35 -> 0.33 | Improved |
| Train with AGM's loss-aligned adversarial objective instead of SSVM training | Structured prediction · Facial emotion intensity prediction, absolute unweighted loss | loss: 0.40 -> 0.33 | Improved |
| Train with AGM's loss-aligned adversarial objective instead of SSVM training | Structured prediction · Facial emotion intensity prediction, squared unweighted loss | loss: 0.40 -> 0.38 | Improved |
| Train with AGM's loss-aligned adversarial objective instead of SSVM training | Structured prediction · Facial emotion intensity prediction, weighted zero-one loss | loss: 0.29 -> 0.28 | Improved |
| Train with AGM's loss-aligned adversarial objective instead of SSVM training | Structured prediction · Facial emotion intensity prediction, weighted absolute loss | loss: 0.29 -> 0.29 | No clear effect |
| Train with AGM's loss-aligned adversarial objective instead of SSVM training | Structured prediction · Facial emotion intensity prediction, weighted squared loss | loss: 0.33 -> 0.36 | Hurt |
| Train with AGM's loss-aligned adversarial objective instead of CRF training | Structured prediction · CoNLL 2005 semantic role labeling | cost-sensitive loss: 0.19 -> 0.14 | Improved |
Closest papers by meaning
- Distributional Robustness with IPMs and links to Regularization and GANs (NeurIPS 2020 · similarity 0.86)
- Adversarial Robustness Guarantees for Gaussian Processes (JMLR 2022 · similarity 0.86)
- Structured Prediction Theory Based on Factor Graph Complexity (NeurIPS 2016 · similarity 0.85)
- Detecting Brittle Decisions for Free: Leveraging Margin Consistency in Deep Robust Classifiers (NeurIPS 2024 · similarity 0.84)
- Alleviating Performance Disparity in Adversarial Spatiotemporal Graph Learning Under Zero-Inflated Distribution (AAAI 2025 · similarity 0.84)
- Fairness-Aware Estimation of Graphical Models (NeurIPS 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
