Research map / Label-efficient and robust learning / Domain adaptation and generalization
An Empirical Investigation of Domain Generalization with Empirical Risk Minimizers
Shanmukha Ramakrishna Vedantam, David Lopez-Paz, David J. Schwab
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 |
|---|---|---|---|
| remove the lambda term from the HdeltaH-MS generalization bound (ablation) | Ben-David bound · DomainBed pooled | Spearman rho with target error: 0.496 -> 0.136 | Hurt |
| remove lambda from the HdeltaH-SS bound (ablation) | Ben-David bound · DomainBed pooled | Spearman rho with target error: 0.509 -> 0.152 | Hurt |
| remove lambda from the H-divergence bound (ablation) | Ben-David bound · DomainBed pooled | Spearman rho with target error: 0.215 -> 0.027 | Hurt |
| use target-domain output entropy as predictor instead of source validation error | OOD accuracy prediction · DomainBed pooled | Spearman rho: 0.712 -> 0.734 | Improved |
| use target entropy instead of source validation error | OOD accuracy prediction · DomainBed pooled | R2: 0.608 -> 0.582 | Hurt |
| use target entropy instead of source validation error | OOD accuracy prediction · RotatedMNIST vs VLCS and PACS | Spearman rho: RotatedMNIST 0.744 -> 0.853; VLCS 0.625 -> 0.557; PACS 0.732 -> 0.662 | Mixed |
| add Entropy-Source to source validation error (joint) | OOD accuracy prediction · DomainBed pooled | Spearman rho: 0.712 -> 0.712 | No clear effect |
| add L2-Path Norm to source validation error (stratified) | OOD accuracy prediction · DomainBed pooled | Spearman rho: 0.712 -> 0.712 | No clear effect |
| add Mixup measure to source validation error (stratified) | OOD accuracy prediction · DomainBed pooled | Spearman rho: 0.712 -> 0.749 | Improved |
| add Fisher Align to source validation error (joint) | OOD accuracy prediction · DomainBed pooled | Spearman rho: 0.441 -> 0.092 | Hurt |
Closest papers by meaning
- Domain Generalization without Excess Empirical Risk (NeurIPS 2022 · similarity 0.91)
- Understanding Domain Generalization: A Noise Robustness Perspective (ICLR 2024 · similarity 0.90)
- Lost Domain Generalization Is a Natural Consequence of Lack of Training Domains (AAAI 2024 · similarity 0.89)
- Robust Domain Generalization under Divergent Marginal and Conditional Distributions (NeurIPS 2026 · similarity 0.88)
- Sharpness-Aware Model-Agnostic Long-Tailed Domain Generalization (AAAI 2024 · similarity 0.88)
- Set Valued Predictions For Robust Domain Generalization (ICML 2025 · similarity 0.88)
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.
