Research map / Label-efficient and robust learning / Recognition and long-tail
On the Minimal Supervision for Training Any Binary Classifier from Only Unlabeled Data
Nan Lu, Gang Niu, Aditya Krishna Menon, Masashi Sugiyama
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 |
|---|---|---|---|
| Train using two unlabeled samples with different class priors and the UU unbiased risk estimator | binary classification · MNIST, Fashion-MNIST, SVHN, and CIFAR-10 | classification error | Improved |
| Use UU instead of pSVM, BER, or BER-FC | binary classification from two unlabeled sets · pendigits, covertype, MNIST, spambase, letter, USPS | classification error | Mixed |
| Move the two training class priors closer together | UU learning and CCN on MNIST · MNIST | classification error | Hurt |
| Use UU rather than CCN as the class-prior gap narrows | learning from two unlabeled sets · MNIST | classification error | Improved |
| Perturb the training class-prior values used by UU | UU · MNIST and CIFAR-10 | classification error: 2.31~(0.14) -> 2.35~(0.14) | No clear effect |
| Use sigmoid loss instead of logistic loss | UU · MNIST | classification error | No clear effect |
| Use BER or BER-FC, which optimize balanced error rather than classification error | two-unlabeled-set learning · datasets with class prior near 0 or 1 | classification error | Hurt |
Closest papers by meaning
- Binary Classification from Multiple Unlabeled Datasets via Surrogate Set Classification (ICML 2021 · similarity 0.87)
- Universal Rates of Empirical Risk Minimization (NeurIPS 2024 · similarity 0.86)
- Positive and Unlabeled Learning with Controlled Probability Boundary Fence (ICML 2024 · similarity 0.86)
- Rethinking Guidance Information to Utilize Unlabeled Samples: A Label Encoding Perspective (ICML 2024 · similarity 0.86)
- AUC Optimization from Multiple Unlabeled Datasets (AAAI 2024 · similarity 0.86)
- On the Provable Advantage of Unsupervised Pretraining (ICLR 2024 · 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.
