Research map / Probabilistic and scientific modeling / Molecules and proteins
Model Shapley: Equitable Model Valuation with Black-box Access
Xinyi Xu, Thanh Lam, Chuan-Sheng Foo, Bryan Kian Hsiang Low
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 |
|---|---|---|---|
| learn on alpha_i features instead of average prediction h_i (ablation) | GPR model appraiser · MNIST, 20% train | MSE: 8.36e−5 -> 1.59e−6 | Improved |
| learn on alpha_i features instead of average prediction h_i (ablation) | GPR model appraiser · MNIST, 20% train | MaxE: 1.59e−2 -> 3.53e−3 | Improved |
| learn on alpha_i features instead of average prediction h_i (ablation) | GPR model appraiser · CIFAR-10, 20% train | MSE: 3.05e−4 -> 1.79e−5 | Improved |
| learn on alpha_i features instead of average prediction h_i (ablation) | GPR model appraiser · CIFAR-10, 20% train | MaxE: 3.23e−2 -> 9.05e−3 | Improved |
| raise training ratio of model-MSV pairs from 5% to 50% | GPR model appraiser · MNIST and CIFAR-10 | test error | Improved |
| use CNN instead of MLP or LR | MSV · MNIST misclassified queries | MSV | Improved |
| multiply top-class probability by 1, 5, 10 (higher predictive certainty) | MSV · MNIST CNN models | MSV | Improved |
| use class-specific Dirichlet abstractions instead of unpartitioned query set (ablation) | MSV · KDD99 imbalanced | consistency with F1 | Improved |
| buy models in highest-MSV-first order instead of lowest-first | random forest of decision trees · Breast Cancer | test accuracy vs number of models | Improved |
| buy models in highest-MSV-first order instead of lowest-first | voting ensemble of LeNets · CIFAR-10 | test accuracy vs number of models | Improved |
Closest papers by meaning
- DU-Shapley: A Shapley Value Proxy for Efficient Dataset Valuation (NeurIPS 2024 · similarity 0.87)
- shapiq: Shapley Interactions for Machine Learning (NeurIPS 2024 · similarity 0.85)
- Generalized Priority-Aware Shapley Value (NeurIPS 2026 · similarity 0.84)
- Data Shapley in One Training Run (ICLR 2025 · similarity 0.83)
- Provably Accurate Shapley Value Estimation via Leverage Score Sampling (ICLR 2025 · similarity 0.83)
- A Sustainable AI Economy Needs Data Deals That Work for Generators (NeurIPS 2025 · similarity 0.83)
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.
