ScholarGate
어시스턴트

방법 비교

선택한 방법을 나란히 검토하세요. 서로 다른 행은 강조 표시됩니다.

강건 능동 학습 (Robust Active Learning)×온라인 학습×
분야머신러닝머신러닝
계열Machine learningMachine learning
기원 연도20061958–2000s
창시자Balcan, M.-F.; Beygelzimer, A.; Langford, J.Rosenblatt, F.; Littlestone, N.; Shalev-Shwartz, S. (key contributors)
유형Active learning with robustness guaranteesLearning paradigm (sequential model update)
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 ↗Shalev-Shwartz, S. (2011). Online Learning and Online Convex Optimization. Foundations and Trends in Machine Learning, 4(2), 107–194. DOI ↗
별칭RAL, noise-tolerant active learning, robust query learning, adversarially robust active learningincremental learning, sequential learning, streaming learning, online machine learning
관련66
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.Online learning is a machine learning paradigm in which a model is updated incrementally as each new data point arrives, rather than being trained once on a fixed dataset. It is essential when data streams continuously, storage is limited, or the underlying distribution shifts over time. Theoretical performance is measured by cumulative regret relative to the best fixed predictor in hindsight.
ScholarGate데이터셋
  1. v1
  2. 2 출처
  3. PUBLISHED
  1. v1
  2. 2 출처
  3. PUBLISHED

검색으로 이동 슬라이드 다운로드

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