← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6698.b · 生成模型:GPT-6

两种词向量目标方向相反

理解等级:中级

CBOW将周围多个词的表示聚合后预测中心词,skip-gram从一个输入词分别预测窗口内邻词。二者均从局部文本构造样本,但CBOW一个窗口与skip-gram多个词对并非相同样本数;不能把用周围词预测缺词的叙述同时当两者定义。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Now that you have your corpus, how exactly do you create a training set for your word-embedding language model? In the early days, there were two main approaches: Continuous bag of words Continuous skip-gram The continuous bag-of-words (CBOW) approach predicts the target word (the output word) from the nearby context words (input words). The only difference between the BOW vectors you learned about in chapter

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第06章《用词嵌入推理》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/chapter-6.html) - 结论依据:摘录中关于两种词向量目标方向相反的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:Now that you have your corpus, how exactly do you create a training set for your word-embedding language model? In the early days, there were two main approaches: Continuous bag of words Continuous skip-gram The continuous bag-of-words (CBOW) approach predicts the target word (the output word) from the nearby context words (input words). The only difference between the BOW vectors you learned about in chapter