Gradient Boosting
Posted RRRR-cord
tags:
篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了Gradient Boosting相关的知识,希望对你有一定的参考价值。
参考网址:
2. Wikipedia: Gradient boosting
一般Gradient Boosting:
输入:训练集$\{(x_{i}, y_{i})\}_{i=1}^{n}$,可导损失函数$L(y, F(x))$,迭代次数$M$
算法:
1. 用常数值初始化模型
\[F_{0}=\argmin_{\gamma} \sum_{i=1}^{n} L(y_{i}, \gamma)\]
2. 从m=1到M:
1)计算“伪残差”:
\[
\gamma_{im}=-[\frac{\partial L(y_{i}, F(x_{i}))}{\partial F(x_{i})}|_{F(x)=F_{m-1}(x)}], i=1, ..., n
\]
\[ \sum_{k=1}^n k^2 = \frac{1}{2} n (n+1).\]
\[ \frac{\partial u}{\partial t}
= h^2 \left( \frac{\partial^2 u}{\partial x^2}
+ \frac{\partial^2 u}{\partial y^2}
+ \frac{\partial^2 u}{\partial z^2}\right)\]
The Newton‘s second law is F=ma.
The Newton‘s second law is $F=ma$.
The Newton‘s second law is
$$F=ma$$
The Newton‘s second law is
\[F=ma\]
Greek Letters $\eta$ and $\mu$
Fraction $\frac{a}{b}$
Power $a^b$
Subscript $a_b$
Derivate $\frac{\partial y}{\partial t} $
Vector $\vec{n}$
Bold $\mathbf{n}$
To time differential $\dot{F}$
Matrix (lcr here means left, center or right for each column)
\[
\left[
\begin{array}{lcr}
a1 & b22 & c333 \\
d444 & e555555 & f6
\end{array}
\right]
\]
Equations(here \& is the symbol for aligning different rows)
\begin{align}
a+b&=c\\
d&=e+f+g
\end{align}
\[
\left\{
\begin{aligned}
&a+b=c\\
&d=e+f+g
\end{aligned}
\right.
\]
以上是关于Gradient Boosting的主要内容,如果未能解决你的问题,请参考以下文章