Research map / Learning theory and algorithms / Matrix and tensor methods
Crowdsourcing via Pairwise Co-occurrences: Identifiability and Algorithms
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 |
|---|---|---|---|
| refine MultiSPA output with KL-divergence fitting (MultiSPA-KL) (ablation) | MultiSPA · UCI Nursery p=0.2 | classification error (%): 17.96 -> 13.06 | Helped |
| refine MultiSPA with KL fitting (ablation) | MultiSPA · synthetic case 2 confusion-matrix estimation, p=0.2 | MSE: 0.0229 -> 0.0029 | Helped |
| refine MultiSPA with KL fitting (ablation) | MultiSPA · AMT TREC | classification error (%): 31.47 -> 29.23 | Helped |
| use second-order MultiSPA-KL instead of third-order TensorADMM | MultiSPA-KL · UCI Adult p=0.2 | classification error (%): 25.08 -> 17.63 | Helped |
| use MultiSPA-KL instead of Minmax-entropy | MultiSPA-KL · AMT Web | classification error (%): 14.58 vs 11.51 | Hurt |
| MultiSPA-KL instead of Minmax-entropy on TREC | MultiSPA-KL · AMT TREC | classification error (%): 91.61 -> 29.23 | Helped |
| initialize EM (Dawid-Skene) with MultiSPA instead of majority voting (ablation) | Dawid-Skene EM · UCI Adult p=0.2 | classification error (%): 75.13 -> 23.88 | Helped |
| unbalanced label prior d=[1/6,2/3,1/6] versus uniform prior | MultiSPA-KL · synthetic case 2, p=0.2 | classification error (%): 31.71 -> 23.19 | Helped |
Closest papers by meaning
- Crowdsourcing via Annotator Co-occurrence Imputation and Provable Symmetric Nonnegative Matrix Factorization (ICML 2021 · similarity 0.92)
- Noisy Label Learning with Instance-Dependent Outliers: Identifiability via Crowd Wisdom (NeurIPS 2024 · similarity 0.86)
- Unbiased Multi-Label Learning from Crowdsourced Annotations (ICML 2024 · similarity 0.86)
- KFNN: K-Free Nearest Neighbor For Crowdsourcing (NeurIPS 2024 · similarity 0.86)
- Learning from Noisy Labels via Conditional Distributionally Robust Optimization (NeurIPS 2024 · similarity 0.84)
- Coupled Confusion Correction: Learning from Crowds with Sparse Annotations (AAAI 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.