Research map / Label-efficient and robust learning / Noisy labels
CS-Shapley: Class-wise Shapley Values for Data Valuation in Classification
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use CS-Shapley class-wise value function instead of TMC-Shapley | TMC-Shapley · Phoneme, Logistic Regression | weighted accuracy drop: 0.009 -> 0.154 | Helped |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · Click, Logistic Regression | weighted accuracy drop: 0.007 -> 0.053 | Helped |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · Diabetes, SVM-RBF | weighted accuracy drop: 0.037 -> 0.106 | Helped |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · Covertype, SVM-RBF | weighted accuracy drop: 0.214 -> 0.193 | Hurt |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · MNIST-10, SVM-RBF | weighted accuracy drop: 0.247 -> 0.203 | Hurt |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · Click, SVM-RBF | weighted accuracy drop: 0.004 -> 0.004 | No clear effect |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · Click, Logistic Regression | noisy label detection AUC: 0.689 -> 0.816 | Helped |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · MNIST-2, SVM-RBF | noisy label detection AUC: 0.663 -> 0.837 | Helped |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · CPU, SVM-RBF | noisy label detection AUC: 0.671 -> 0.808 | Helped |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · Covertype, Logistic Regression | noisy label detection AUC: 0.766 -> 0.706 | Hurt |
| use CS-Shapley instead of TMC-Shapley | TMC-Shapley · MNIST-10, Logistic Regression | noisy label detection AUC: 0.933 -> 0.877 | Hurt |
| use CS-Shapley instead of Leave-One-Out | Leave-One-Out · CPU, Logistic Regression | weighted accuracy drop: 0.040 -> 0.036 | Hurt |
Closest papers by meaning
- DU-Shapley: A Shapley Value Proxy for Efficient Dataset Valuation (NeurIPS 2024 · similarity 0.89)
- Rethinking Data Shapley for Data Selection Tasks: Misleads and Merits (ICML 2024 · similarity 0.88)
- A Unified Framework for Provably Efficient Algorithms to Estimate Shapley Values (NeurIPS 2025 · similarity 0.87)
- shapiq: Shapley Interactions for Machine Learning (NeurIPS 2024 · similarity 0.86)
- Amortized Linear-time Exact Shapley Value for Product-Kernel Methods (NeurIPS 2026 · similarity 0.85)
- Shapley-Guided Utility Learning for Effective Graph Inference Data Valuation (ICLR 2025 · similarity 0.85)
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.