陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记

深入研究了一下陈天奇Boosted Tree的PPT,做了点简单的笔记,可以说是PPT的缩略版:
框架有了,截了些重要的图和公式。
虽然简略,但是足以学习大牛思考问题的方式。

Review of key concepts of supervised learning

  • Elements in Supervised Learning
    • Model
    • Parameters
    • Objective function
  • Putting known knowledge into context
  • Objective and Bias Variance Trade-off
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第1张图片

Regression Tree and Ensemble (What are we Learning)

  • Regression Tree (CART)
  • Regression Tree Ensemble
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第2张图片
  • Tree Ensemble methods(基于树的集成方法的一些优点)
  • Put into context: Model and Parameters(model:加法模型;parameters:树/函数)
  • Learning a tree on single variable(情歌率-时间序列的例子)
  • Learning a step function(阶跃函数)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第3张图片
  • Learning step function (visually)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第4张图片
  • Coming back: Objective for Tree Ensemble
  • Objective vs Heuristic
  • Regression Tree is not just for regression!

Gradient Boosting (How do we Learn)

  • Take Home Message for this section (其实是总结第二部分。。)
  • So How do we Learn? (SGD不可以,提出加法模型)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第5张图片
  • Additive Training (公式推导,生成残差)
  • Taylor Expansion Approximation of Loss (二级泰勒展开,出现gi和hi)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第6张图片
  • Our New Goal (化简后的损失函数)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第7张图片
  • Refine the definition of tree (树的数学表达)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第8张图片
  • Define Complexity of a Tree (cont’) (定义树的复杂度)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第9张图片
  • Revisit the Objectives (结合前两张,重看目标函数)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第10张图片
  • The Structure Score
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第11张图片
  • The Structure Score Calculation
  • Searching Algorithm for Single Tree
  • Greedy Learning of the Tree
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第12张图片
  • Efficient Finding of the Best Split
  • An Algorithm for Split Finding
  • What about Categorical Variables?
  • Pruning and Regularization
  • Recap: Boosted Tree Algorithm (总结第三部分)
陈天奇《Introduction to Boosted Trees》PPT 缩略版笔记_第13张图片

你可能感兴趣的:(机器学习)