Research map / Continual learning / Online and replay-based
Active Lifelong Learning With "Watchdog"
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 |
|---|---|---|---|
| replace random task order with watchdog reconstruction-cost selection | ELLA · Landmine | AUC: 76.003 -> 78.190 | Helped |
| replace random task order with watchdog selection | ELLA · School | RMSE: 10.129 -> 9.982 | Helped |
| replace information-maximization selection with watchdog selection | ELLA · Yeast | AUC: 68.230 -> 68.238 | No clear effect |
| replace ELLA-Diver++ with watchdog selection on first half of tasks | ELLA · Yeast | AUC: 63.43 -> 62.76 | Hurt |
| replace ELLA-Info with watchdog selection | ELLA · Parkinson-Total | RMSE: 2.448 -> 2.421 | Helped |
| use ELLA-based task selection instead of self-paced spMTFL | ELLA · Yeast | AUC: 58.720 -> 68.238 | Helped |
| use hierarchical two-level dictionary instead of single-level dictionary methods | dictionary learning · Extended Yale B | accuracy (%): 95.70 -> 95.89 | Helped |
| add watchdog selection on top of ELLA (runtime cost) | ELLA · Landmine | runtime (s): 0.97 -> 1.50 | Hurt |
| vary lambda2 of local dictionary (ablation) | AcLL dictionary · Extended Yale B | accuracy (%) | Mixed |
Closest papers by meaning
- Task-aware Orthogonal Sparse Network for Exploring Shared Knowledge in Continual Learning (ICML 2024 · similarity 0.84)
- Model Sensitivity Aware Continual Learning (NeurIPS 2024 · similarity 0.83)
- Beyond task diversity: provable representation transfer for sequential multitask linear bandits (NeurIPS 2024 · similarity 0.83)
- Privacy-Aware Lifelong Learning (ICLR 2025 · similarity 0.83)
- Mitigating Catastrophic Forgetting in Online Continual Learning by Modeling Previous Task Interrelations via Pareto Optimization (ICML 2024 · similarity 0.82)
- Adaptive Discovering and Merging for Incremental Novel Class Discovery (AAAI 2024 · similarity 0.82)
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.