השוואת שיטות

	LSTM מפוקח-בקושי	LSTM מכוונן עדין
תחום	למידה עמוקה	למידה עמוקה
משפחה	Machine learning	Machine learning
שנת המקור	2016–2018	2018 (fine-tuning paradigm formalised); LSTM core: 1997
הוגה השיטה	Ratner et al. (data programming framework); Hochreiter & Schmidhuber (LSTM backbone)	Howard, J. & Ruder, S. (ULMFiT); foundational LSTM by Hochreiter & Schmidhuber
סוג	Weakly supervised sequence model	Supervised sequential model with transfer learning
מקור מכונן	Ratner, A., De Sa, C., Wu, S., Selsam, D., & Re, C. (2016). Data Programming: Creating Large Training Sets, Quickly. Advances in Neural Information Processing Systems (NeurIPS), 29.	Howard, J., & Ruder, S. (2018). Universal Language Model Fine-tuning for Text Classification. Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (ACL), 328–339.
כינויים	WS-LSTM, noisy-label LSTM, distant-supervision LSTM, data-programming LSTM	Fine-Tuned LSTM, LSTM Fine-Tuning, Pre-trained LSTM with Task Adaptation, LSTM Transfer Learning
קשורות	6	6
תקציר	Weakly supervised LSTM trains a Long Short-Term Memory network on sequence data where clean, manually annotated labels are scarce or absent. Instead, multiple imperfect label sources — heuristic rules, distant supervision, crowdsourcing, or programmatic labeling functions — are combined to produce probabilistic training labels, which are then used to supervise the LSTM. This allows scalable training on large unlabeled corpora without exhaustive human annotation.	Fine-Tuned LSTM adapts a Long Short-Term Memory network pre-trained on a large corpus to a specific downstream task — such as text classification, sentiment analysis, or sequence labeling — by continuing training on task-specific labeled data. Popularised by the ULMFiT framework, this approach achieves strong performance even when labeled data is scarce.