ScholarGate
어시스턴트

방법 비교

선택한 방법을 나란히 검토하세요. 서로 다른 행은 강조 표시됩니다.

강건 능동 학습 (Robust Active Learning)×퓨샷 학습×
분야머신러닝머신러닝
계열Machine learningMachine learning
기원 연도20062011–2017
창시자Balcan, M.-F.; Beygelzimer, A.; Langford, J.Lake, B. M.; Vinyals, O.; Finn, C. et al.
유형Active learning with robustness guaranteesMeta-learning / low-data learning paradigm
원전Balcan, M.-F., Beygelzimer, A., & Langford, J. (2006). Agnostic active learning. In Proceedings of the 23rd International Conference on Machine Learning (ICML 2006), pp. 65–72. ACM. 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 ↗
별칭RAL, noise-tolerant active learning, robust query learning, adversarially robust active learningFSL, low-shot learning, k-shot learning, meta-learning for few examples
관련64
요약Robust Active Learning extends the standard active learning framework to handle noisy labels, adversarial perturbations, and unreliable oracles. Rather than assuming perfect labeling, it incorporates statistical or adversarial robustness guarantees into the query selection process, maintaining sample efficiency while tolerating corruption in the annotation process.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.
ScholarGate데이터셋
  1. v1
  2. 2 출처
  3. PUBLISHED
  1. v1
  2. 2 출처
  3. PUBLISHED

검색으로 이동 슬라이드 다운로드

ScholarGate방법 비교: Robust Active Learning · Few-shot Learning. 2026-06-15에 다음에서 검색함: https://scholargate.app/ko/compare