Research map / Label-efficient and robust learning / Recognition and long-tail
Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach
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 probabilistic logic for classifier error estimation and label combination | LEE · NELL-7 | Error MAD: 0.229 -> 0.152 | Helped |
| Use probabilistic logic for classifier error estimation and label combination | LEE · NELL-11 | Target AUC: 0.477 -> 0.615 | Helped |
| Use probabilistic logic for classifier error estimation and label combination | LEE · uNELL-All | Error MAD: 0.42 -> 0.37 | Helped |
| Use probabilistic logic for classifier error estimation and label combination | LEE · uNELL-All | Target AUC: 99.9 -> 96.5 | Hurt |
| Use probabilistic logic for classifier error estimation and label combination | LEE · uNELL-10% | Error MAD: 1.90 -> 0.43 | Helped |
| Use probabilistic logic for classifier error estimation and label combination | LEE · uBRAIN-All | Error MAD: 0.40 -> 0.38 | Helped |
| Use probabilistic logic for classifier error estimation and label combination | LEE · uBRAIN-10% | Error MAD: 0.63 -> 0.47 | Helped |
Closest papers by meaning
- Estimating Accuracy from Unlabeled Data: A Bayesian Approach (ICML 2016 · similarity 0.95)
- Recovering True Classifier Performance in Positive-Unlabeled Learning (AAAI 2017 · similarity 0.87)
- Leveraging Uncertainty Estimates To Improve Classifier Performance (ICLR 2024 · similarity 0.85)
- A Data-Driven Measure of Relative Uncertainty for Misclassification Detection (ICLR 2024 · similarity 0.84)
- Don’t Label Twice: Quantity Beats Quality when Comparing Binary Classifiers on a Budget (ICML 2024 · similarity 0.84)
- Naive Bayes Classifiers over Missing Data: Decision and Poisoning (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.