Research map / Label-efficient and robust learning / Recognition and long-tail
Improved Bayes Risk Can Yield Reduced Social Welfare Under Competition
Meena Jagadeesan, Michael Jordan, Jacob Steinhardt, Nika Haghtalab
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 |
|---|---|---|---|
| Improve the representation from the trivial to the informative representation | Competing predictors · Binary classification, simple mathematical example | equilibrium social loss: 0 -> 0.1 | Hurt |
| Improve representation quality by varying per-representation Bayes risk | Competing predictors · Stylized binary classification, single representation | equilibrium social loss | Mixed |
| Improve representation quality by reducing representation noise | Competing predictors · Stylized binary classification, Gaussian mixture | equilibrium social loss | Mixed |
| Improve representation quality by increasing representation dimension | Competing predictors · Stylized binary classification, synthetic data | equilibrium social loss | Mixed |
| Improve representation quality by varying per-representation Bayes risk | Linear predictors with noisy user decisions · Synthetic binary classification | equilibrium social loss | Mixed |
| Improve representation quality by reducing representation noise | Linear predictors with noisy user decisions · Synthetic binary classification, Gaussian mixture | equilibrium social loss | Mixed |
| Improve representation quality by increasing representation dimension | Linear predictors with noisy user decisions · Synthetic binary classification | equilibrium social loss | Mixed |
| Use VGG16 rather than AlexNet representations | Competing linear predictors · CIFAR-10 binary classification, m = 3 | equilibrium social loss | Mixed |
| Use AlexNet rather than ResNet50 representations | Competing linear predictors · CIFAR-10 binary classification, m = 5 and m = 8 | equilibrium social loss | Mixed |
| Use ResNet50 rather than AlexNet representations | Competing linear predictors · CIFAR-10 binary classification, m = 3 and m = 4 | equilibrium social loss | Mixed |
| Use ResNet18 rather than VGG16 representations | Competing linear predictors · CIFAR-10 10-class classification | equilibrium social loss | Mixed |
| Use VGG16 rather than AlexNet representations | Competing linear predictors · CIFAR-10 10-class classification, m = 3 | equilibrium social loss | Mixed |
Closest papers by meaning
- A Market for Accuracy: Classification Under Competition (ICML 2025 · similarity 0.87)
- Accuracy vs. Accuracy: Computational Tradeoffs Between Classification Rates and Utility (NeurIPS 2026 · similarity 0.86)
- The Hidden Cost of Waiting for Accurate Predictions (ICLR 2025 · similarity 0.86)
- Decongestion by Representation: Learning to Improve Economic Welfare in Marketplaces (ICLR 2024 · similarity 0.85)
- Model Alignment as Prospect Theoretic Optimization (ICML 2024 · similarity 0.84)
- The Relative Value of Prediction in Algorithmic Decision Making (ICML 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.
