일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | ||||
4 | 5 | 6 | 7 | 8 | 9 | 10 |
11 | 12 | 13 | 14 | 15 | 16 | 17 |
18 | 19 | 20 | 21 | 22 | 23 | 24 |
25 | 26 | 27 | 28 | 29 | 30 | 31 |
- Gradient descent
- tensorflow
- 폴댄스
- 자연어처리
- 파이썬과 케라스로 배우는 강화학습
- MDP
- model free
- 머신러닝
- 애쉬브라운
- 다이내믹 프로그래밍
- 딥러닝
- NLP
- 인공지능
- 취미
- Linear Regression
- CS224N
- MB
- 리처드소처
- 딥뉴럴넷
- 강화학습
- 파이썬
- Richard Socher
- 행복
- cost function
- 벨만 방정식
- 독서
- kbpf
- 페미니즘
- 양성평등
- ai
- Today
- Total
목록Study/Paper (2)
Ordinary day

논문 출처 https://arxiv.org/abs/1905.01969 Poly-encoders: Transformer Architectures and Pre-training Strategies for Fast and Accurate Multi-sentence Scoring The use of deep pre-trained bidirectional transformers has led to remarkable progress in a number of applications (Devlin et al., 2018). For tasks that make pairwise comparisons between sequences, matching a given input with a corresponding labe..
Learning from Demonstrations for Real World Reinforcement Learning - https://arxiv.org/abs/1704.03732실제 상황에서의 문제점이 발생따라서 실제 데이터로 학습이 필요함그래서 제안한 DQfD *background2015년에 나온 더블DQN: 타겟 네트워크 분리, replay Buffer2016년 : 현재 예측하는 네트워크와 타겟 네트워크 2개 존재 더블DQN 로스 사용2016년(2): DQN 에이전트 수정 *related work모방학습 DAGGER(2011)zero sum game알파고Human Experience Replay: 샘플링 방법은 같으나, supervised loss사용 안함Replay Buffer Spiking: ..