Don't Look Back in Anger: MAGIC Net for Streaming Continual Learning with Temporal Dependence
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866918380220448768 |
|---|---|
| author | Giannini, Federico D'Andrea, Sandro Della Valle, Emanuele |
| author_facet | Giannini, Federico D'Andrea, Sandro Della Valle, Emanuele |
| contents | Concept drift, temporal dependence, and catastrophic forgetting represent major challenges when learning from data streams. While Streaming Machine Learning and Continual Learning (CL) address these issues separately, recent efforts in Streaming Continual Learning (SCL) aim to unify them. In this work, we introduce MAGIC Net, a novel SCL approach that integrates CL-inspired architectural strategies with recurrent neural networks to tame temporal dependence. MAGIC Net continuously learns, looks back at past knowledge by applying learnable masks over frozen weights, and expands its architecture when necessary. It performs all operations online, ensuring inference availability at all times. Experiments on synthetic and real-world streams show that it improves adaptation to new concepts, limits memory usage, and mitigates forgetting. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_08600 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Don't Look Back in Anger: MAGIC Net for Streaming Continual Learning with Temporal Dependence Giannini, Federico D'Andrea, Sandro Della Valle, Emanuele Machine Learning Artificial Intelligence 68T05, 68T07 I.2.6; I.2.7 Concept drift, temporal dependence, and catastrophic forgetting represent major challenges when learning from data streams. While Streaming Machine Learning and Continual Learning (CL) address these issues separately, recent efforts in Streaming Continual Learning (SCL) aim to unify them. In this work, we introduce MAGIC Net, a novel SCL approach that integrates CL-inspired architectural strategies with recurrent neural networks to tame temporal dependence. MAGIC Net continuously learns, looks back at past knowledge by applying learnable masks over frozen weights, and expands its architecture when necessary. It performs all operations online, ensuring inference availability at all times. Experiments on synthetic and real-world streams show that it improves adaptation to new concepts, limits memory usage, and mitigates forgetting. |
| title | Don't Look Back in Anger: MAGIC Net for Streaming Continual Learning with Temporal Dependence |
| topic | Machine Learning Artificial Intelligence 68T05, 68T07 I.2.6; I.2.7 |
| url | https://arxiv.org/abs/2603.08600 |