Skip to content

Commit

Permalink
chapter 5 update
Browse files Browse the repository at this point in the history
  • Loading branch information
liber145 committed Jun 8, 2017
1 parent 6179e86 commit 60de433
Show file tree
Hide file tree
Showing 2 changed files with 5 additions and 2 deletions.
4 changes: 2 additions & 2 deletions Chapter5/machine_learning_basics.tex
Original file line number Diff line number Diff line change
Expand Up @@ -1088,7 +1088,7 @@ \subsection{条件对数似然和\glsentrytext{mean_squared_error}}
=& -m \log\sigma - \frac{m}{2} \log(2\pi) - \sum_{i=1}^m \frac{ \norm{\hat{y}^{(i)} - y^{(i)} }^2 }{2\sigma^2},
\end{align}
其中$\hat{y}^{(i)}$\gls{linear_regression}在第$i$个输入$\Vx^{(i)}$上的输出,$m$是训练\gls{example:chap5}的数目。
对比于\gls{mean_squared_error}的对数似然
对比\gls{mean_squared_error}和对数似然
\begin{equation}
\text{MSE}_{\text{train}} = \frac{1}{m} \sum_{i=1}^m \norm{\hat{y}^{(i)} - y^{(i)}}^2,
\end{equation}
Expand Down Expand Up @@ -1246,7 +1246,7 @@ \section{贝叶斯统计}
-\frac{1}{2} (\Vw - \Vmu_m)^\Tsp \VLambda_m^{-1} (\Vw - \Vmu_m)
\right).
\end{align}
所有不包括的参数向量$\Vw$的项都已经被删去了;它们意味着分布的积分必须归一这个事实
分布的积分必须归一这个事实意味着要删去所有不包括参数向量$\Vw$的项
\eqnref{eq:3.23}显示了如何标准化多元高斯分布。

% -- 134 --
Expand Down
3 changes: 3 additions & 0 deletions acknowledgments_github.md
Original file line number Diff line number Diff line change
Expand Up @@ -221,3 +221,6 @@
---------------
- @neo0801 ==> Chapter2, 2.2节一处翻译调整

2017年6月7日
---------------
- @tjliupeng ==> Chapter 5, 翻译调整

0 comments on commit 60de433

Please sign in to comment.