Sammenlign metoder
Gennemgå dine valgte metoder side om side; rækker, der afviger, er fremhævet.
| Overførselslæring× | Few-shot Learning× | |
|---|---|---|
| Fagområde | Maskinlæring | Maskinlæring |
| Familie | Machine learning | Machine learning |
| Oprindelsesår≠ | 2010 (formalized); 1990s (early roots) | 2011–2017 |
| Ophavsperson≠ | Pan, S. J. & Yang, Q. (survey); Bengio, Y. (deep learning framing) | Lake, B. M.; Vinyals, O.; Finn, C. et al. |
| Type≠ | Learning paradigm | Meta-learning / low-data learning paradigm |
| Oprindelig kilde≠ | Pan, S. J., & Yang, Q. (2010). A Survey on Transfer Learning. IEEE Transactions on Knowledge and Data Engineering, 22(10), 1345–1359. DOI ↗ | Vinyals, O., Blundell, C., Lillicrap, T., Wierstra, D., & Kavukcuoglu, K. (2016). Matching Networks for One Shot Learning. Advances in Neural Information Processing Systems (NeurIPS), 29. link ↗ |
| Aliasser | TL, domain adaptation, fine-tuning, pre-trained model adaptation | FSL, low-shot learning, k-shot learning, meta-learning for few examples |
| Relaterede≠ | 3 | 4 |
| Resumé≠ | Transfer learning is a machine learning paradigm in which knowledge gained from training a model on a source task or domain is reused to improve learning on a different but related target task or domain. It is especially powerful when labeled data for the target task is scarce, and it underlies most modern deep learning applications in computer vision, natural language processing, and beyond. | Few-shot learning is a machine learning paradigm that trains models to recognize new classes or solve new tasks from only a handful of labeled examples — typically one to five — by leveraging prior knowledge acquired from a large, related training distribution. It is especially relevant in domains where labeling is expensive, scarce, or structurally limited. |
| ScholarGateDatasæt ↗ |
|
|