Skip to content

Instantly share code, notes, and snippets.

@magigo

magigo/上下文定向.md

Created February 4, 2015 10:07
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save magigo/9fabfd13c24e7e903d02 to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save magigo/9fabfd13c24e7e903d02 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
上下文定向.md

  1. cache: url -> 特征表

  2. 话题模型 PLSI，LDA，GaP 经验贝叶斯

    LDA可以视为PLSI的经验贝叶斯版本

    PLSI不是指数族分布，不能使用EM算法

    可用变分近似，叫做VBEM简单，但无法保证收敛到局部最优

    概率方法，以概率1收敛到局部最优

    并行化E-step(mapper), M-step(reducer)更新

  3. 没有免费的午餐定理，没有先验的假设任何模型的平均性能都是一样的

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment