Research map / Continual learning / Online and replay-based
Online Continual Learning from Imbalanced Data
Aristotelis Chrysakis, Marie-Francine Moens
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 |
|---|---|---|---|
| Replace reservoir sampling / GSS memory population with class-balancing reservoir sampling (CBRS) | online continual learning memory population · MNIST, Fashion-MNIST, CIFAR-10, CIFAR-100 (imbalanced streams) | test accuracy: MNIST 83.3 vs 74.2 (GSS); CIFAR-100 40.2 vs 28.1 (reservoir) | Improved |
| Use uniform instead of weighted replay with CBRS (ablation) | CBRS · MNIST m=1000 and CIFAR-10 m=1000 | test accuracy: MNIST 86.0 vs 86.1; CIFAR-10 74.4 vs 73.7 (m=1000) | No clear effect |
| Use weighted instead of uniform replay with reservoir sampling (ablation) | RESERVOIR · MNIST m=1000 and CIFAR-10 m=1000 | test accuracy: MNIST 76.3 vs 71.3; CIFAR-10 56.7 vs 54.2 | Improved |
| Use weighted instead of uniform replay with CBRS at m=5000 (ablation) | CBRS · MNIST and CIFAR-10 m=5000 | test accuracy: MNIST 90.3 vs 85.6; CIFAR-10 78.3 vs 76.8 | Mixed |
| Increase memory size from 100 to 5000 with CBRS | CBRS · Fashion-MNIST | test accuracy: 66.5 at m=100 to 77.1 at m=5000 | Improved |
| Add CBRS memory overhead (index structure per class) vs GSS gradient storage | memory population · CIFAR-100 compute | wall-clock time per batch and memory overhead: CBRS 0.03% overhead vs GSS 3639.86%; CBRS 3671.1 ms vs GSS 364.9 ms (time per batch) | Improved |
| Replace reservoir sampling with GSS-Greedy memory population | online continual learning memory population · CIFAR-100 (ResNet-18, m=1000) | test accuracy: 23.0 vs 28.1 for reservoir | Hurt |
| Use weighted replay (inverse-class-count sampling) instead of uniform replay for all methods (ablation) | online CL replay · MNIST, CIFAR-10 at m=1000 and m=5000 | test accuracy: Gains for RESERVOIR/GSS; none for CBRS at m=1000 (86.0 vs 86.1) | Mixed |
Closest papers by meaning
- Gradient Reweighting: Towards Imbalanced Class-Incremental Learning (CVPR 2024 · similarity 0.87)
- Summarizing Stream Data for Memory-Constrained Online Continual Learning (AAAI 2024 · similarity 0.87)
- Defying Imbalanced Forgetting in Class Incremental Learning (AAAI 2024 · similarity 0.87)
- Predicting the Susceptibility of Examples to Catastrophic Forgetting (ICML 2025 · similarity 0.87)
- Do Your Best and Get Enough Rest for Continual Learning (CVPR 2025 · similarity 0.87)
- Online Task-Free Continual Learning via Dynamic Expansionable Memory Distribution (CVPR 2025 · similarity 0.87)
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.
