Research map / Optimization / Nonconvex and smooth optimization
Balanced Self-Paced Learning for AUC Maximization
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 |
|---|---|---|---|
| Add self-paced sample weighting to the kernel-based AUC method | BSPAUC (kernel) · Original benchmark datasets | AUC | Helped |
| Add self-paced sample weighting to the deep AUC method | BSPAUC (deep) · Original benchmark datasets | AUC | Helped |
| Use BSPAUC instead of existing AUC maximization baselines | BSPAUC · Noisy benchmark datasets (label flipping and poison injection, 10%–30%) | AUC | Helped |
| Increase the maximum pace parameter λ∞ (ablation) | BSPAUC (deep) · Datasets with 20% injected poison samples | AUC | Mixed |
| Increase the balance regularization weight μ (ablation) | BSPAUC · Datasets with 20% injected poison samples | Absolute proportion difference (APD) | Helped |
| Add balanced self-paced regularization to balance selected positive and negative proportions (ablation) | BSPAUC · AUC maximization with imbalanced classes | AUC | Helped |
| Exclude high-loss samples by assigning them zero weight during self-paced training | BSPAUC · Noisy benchmark datasets | AUC | Helped |
Closest papers by meaning
- Stochastic AUC Maximization with Deep Neural Networks (ICLR 2020 · similarity 0.87)
- Self-Paced Learning: An Implicit Regularization Perspective (AAAI 2017 · similarity 0.87)
- Fast Stochastic AUC Maximization with $O(1/n)$-Convergence Rate (ICML 2018 · similarity 0.87)
- Pairwise AUC Optimization Needs Corrective Power: A Unified View (NeurIPS 2026 · similarity 0.86)
- A Federated Stochastic Multi-level Compositional Minimax Algorithm for Deep AUC Maximization (ICML 2024 · similarity 0.85)
- Symmetric Self-Paced Learning for Domain Generalization (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.