Uporedite metode
Pregledajte izabrane metode jednu pored druge; redovi koji se razlikuju su istaknuti.
| Polu-nadgledani Word2Vec× | Polu-nadgledana BERT-bazirana klasifikacija× | |
|---|---|---|
| Oblast | Duboko učenje | Duboko učenje |
| Porodica | Machine learning | Machine learning |
| Godina nastanka≠ | 2013–2015 | 2019–2020 |
| Tvorac≠ | Mikolov, T. et al. (Word2Vec); semi-supervised framing via Collobert & Weston and subsequent NLP literature | Multiple groups (Xie et al.; Chen et al.; Devlin et al. for BERT base) |
| Tip≠ | Semi-supervised representation learning | Semi-supervised fine-tuning of pre-trained transformer |
| Temeljni izvor≠ | Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013). Efficient Estimation of Word Representations in Vector Space. In Proceedings of ICLR 2013. link ↗ | Xie, Q., Dai, Z., Hovy, E., Luong, T., & Le, Q. (2020). Unsupervised Data Augmentation for Consistency Training. Advances in Neural Information Processing Systems (NeurIPS), 33, 27780–27792. link ↗ |
| Drugi nazivi | Word2Vec with semi-supervised learning, semi-supervised word embeddings, Word2Vec SSL, unsupervised pretraining with Word2Vec | Semi-supervised BERT, BERT SSL Classification, BERT with Unlabeled Data, BERT Semi-supervised Fine-tuning |
| Srodne | 6 | 6 |
| Sažetak≠ | Semi-supervised Word2Vec trains dense word representations on a large unlabeled corpus using Word2Vec (skip-gram or CBOW), then uses those embeddings as fixed or fine-tunable input features for a downstream classifier trained on a small labeled dataset. This two-stage process lets models benefit from abundant unlabeled text when labeled data is scarce. | Semi-supervised BERT-based classification fine-tunes a pre-trained BERT encoder on a small pool of labeled text examples while simultaneously leveraging a much larger body of unlabeled text — via consistency training, pseudo-labeling, or data augmentation — to produce high-quality classifiers even when manual annotation is scarce. |
| ScholarGateSkup podataka ↗ |
|
|