Skip to content

Instantly share code, notes, and snippets.

@magigo

magigo/点击率预测与逻辑回归.md

Created February 5, 2015 05:07
Show Gist options
  • Save magigo/b2f4e2bdf0af83111343 to your computer and use it in GitHub Desktop.
Save magigo/b2f4e2bdf0af83111343 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
点击率预测与逻辑回归.md
  1. p(click | a,u,c)
  2. 回归(Regression)比排序(Ranking)更合适
  3. cold-start，利用广告层级结构(creative,solution,campaign,advertiser)，以及广告标签对新广告点击率做预测
  4. 捕获点击率的动态特性，快速调整特征，快速调整模型

视角1，逻辑回归是广义线性模型在Binomial error情形的特例

视角2，最大熵模型在类数目等于2的情形特例

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