Starry, starry night

Beijing, China IIIS@Tsinghua @strongman1995
  1. Relation Extraction

    Relation Extraction

    2019/08/12

  2. Transformer

    Transformer

    2019/08/12

  3. pre-train

    从现在的大趋势来看,使用某种模型预训练一个语言模型看起来是一种比较靠谱的方法。从之前AI2的 ELMo,到 OpenAI的fine-tune transformer,再到Google的这个BERT,全都是对预训练的语言模型的应用。

    2019/08/10

  4. weakly supervision

    弱监督学习

    2019/08/09

  5. Attention

    Attention

    2019/08/05

  6. 分类器性能指标:ROC曲线、AUC值

    ROC 曲线

    2019/08/01

  7. 梯度消失和梯度爆炸

    梯度消失、爆炸

    2019/07/28

  8. Reinforcement Learning

    Markov Decision Process(MDP)

    2019/07/21

  9. Adversarial Learning

    traditional ML:optimization: Adversarial ML: game theory:

    2019/07/21

  10. Attention

    Attention Mechanism

    2019/07/19

自动作诗

0 0

strongman's blog website

0 0

homework for statistical learning theory

0 0

homework for computational linguistics

0 0

a company information website

0 0