Research map / Label-efficient and robust learning / Recognition and long-tail
Storage Fit Learning with Feature Evolvable Streams
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 manifold regularization to online feature-evolvable learners (ablation) | NOGD, uROGD, and fROGD · Credit-a, Diabetes, Svmguide3, Swiss, RFID, HTRU_2, magic04 | accuracy: NOGD: .690±.051 -> .706±.049 (Credit-a); uROGD: .658±.021 -> .678±.015 (Diabetes); fROGD: .702±.073 -> .811±.067 (Swiss) | Helped |
| Combine the recovered-space and new-space learners with risk-based weights, using MR and buffer size 60 | SF²EL · Credit-a, Diabetes, Svmguide3, Swiss, RFID, HTRU_2, magic04 | accuracy: Credit-a: .759±.028 -> .768±.029; Diabetes: .678±.015 -> .685±.011; Svmguide3: .686±.039 -> .694±.040; Swiss: .855±.034 | Helped |
| Increase buffer size from 10 to 60 (ablation) | SF²EL · Credit-a, Diabetes, Svmguide3, Swiss, RFID, HTRU_2, magic04 | accuracy: Credit-a: .659±.052 -> .768±.029; Diabetes: .631±.066 -> .685±.011; Svmguide3: .644±.103 -> .694±.040; Swiss: .290±.070 | Helped |
| Increase buffer size from 20 to 60 (ablation) | SF²EL · Credit-a, Diabetes, Svmguide3, Swiss, RFID, HTRU_2, magic04 | accuracy: Credit-a: .737±.036 -> .768±.029; Diabetes: .666±.025 -> .685±.011; Svmguide3: .655±.064 -> .694±.040; Swiss: .617±.067 | Helped |
| Increase buffer size from 40 to 60 (ablation) | SF²EL · Credit-a, Diabetes, Svmguide3, Swiss, RFID, HTRU_2, magic04 | accuracy: Credit-a: .755±.039 -> .768±.029; Diabetes: .676±.016 -> .685±.011; Svmguide3: .683±.034 -> .694±.040; Swiss: .861±.031 | Helped |
Closest papers by meaning
- One-Pass Feature Evolvable Learning with Theoretical Guarantees (ICML 2025 · similarity 0.87)
- Fast Co-Training under Weak Dependence via Stream-Based Active Learning (ICML 2024 · similarity 0.82)
- Elastic Feature Consolidation For Cold Start Exemplar-Free Incremental Learning (ICLR 2024 · similarity 0.82)
- Test-time Adaptation in Non-stationary Environments via Adaptive Representation Alignment (NeurIPS 2024 · similarity 0.81)
- Online Boosting Adaptive Learning under Concept Drift for Multistream Classification (AAAI 2024 · similarity 0.81)
- Data-Free Reservoir Features for Efficient Long-Horizon Cold-Start Continual Learning (NeurIPS 2026 · similarity 0.81)
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.