Research map / Learning theory and algorithms / Clustering
Analysis of Minimax Error Rate for Crowdsourcing and Its Application to Worker Clustering Model
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 worker clustering with L = 10 instead of DS | Crowdsourcing ground-truth estimation · RTE | error rate (%): 7.62 -> 6.75 | Helped |
| increase the maximum cluster count from L = 1 to L = 10 (ablation) | Worker clustering · RTE | error rate (%): 11.37 -> 6.75 | Helped |
| choose L to minimize R(ρ, π) rather than use L = 10 (ablation) | Worker clustering · RTE | error rate (%): 6.75 -> 6.75 | No clear effect |
| use worker clustering with L = 10 instead of DS | Crowdsourcing ground-truth estimation · Bird | error rate (%): 15.74 -> 11.11 | Helped |
| increase the maximum cluster count from L = 1 to L = 10 (ablation) | Worker clustering · Bird | error rate (%): 19.44 -> 11.11 | Helped |
| choose L to minimize R(ρ, π) rather than use L = 10 (ablation) | Worker clustering · Bird | error rate (%): 11.11 -> 11.11 | No clear effect |
| use worker clustering with L chosen by minimizing R(ρ, π) instead of DS | Crowdsourcing ground-truth estimation · Dog | error rate (%): 16.60 -> 16.36 | Helped |
| increase the maximum cluster count from L = 1 to L = 10 (ablation) | Worker clustering · Dog | error rate (%): 18.71 -> 16.98 | Helped |
| choose L to minimize R(ρ, π) rather than use L = 10 (ablation) | Worker clustering · Dog | error rate (%): 16.98 -> 16.36 | Helped |
| cluster workers to pool labels from workers with few annotations | Worker clustering versus DS and MV · Synthetic ZLN setting | accuracy | Helped |
| increase the maximum number of worker clusters L (ablation) | Worker clustering · Synthetic and real-world datasets | empirical error | Helped |
Closest papers by meaning
- A Minimax Optimal Algorithm for Crowdsourcing (NeurIPS 2017 · similarity 0.91)
- Exact Exponent in Optimal Rates for Crowdsourcing (ICML 2016 · similarity 0.90)
- Exploiting Worker Correlation for Label Aggregation in Crowdsourcing (ICML 2019 · similarity 0.90)
- KFNN: K-Free Nearest Neighbor For Crowdsourcing (NeurIPS 2024 · similarity 0.85)
- Mixture of Experts Based Multi-Task Supervise Learning from Crowds (AAAI 2025 · similarity 0.84)
- Label Distribution Propagation-based Label Completion for Crowdsourcing (ICML 2025 · 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.