coursera 前三张公式速查

最新推荐文章于 2020-10-14 08:30:00 发布

tjl_moby

最新推荐文章于 2020-10-14 08:30:00 发布

阅读量236

点赞数

分类专栏： coursera ml笔记 coursera笔记

本文链接：https://blog.csdn.net/tjl_moby/article/details/65648134

版权

coursera笔记同时被 2 个专栏收录

17 篇文章 0 订阅

订阅专栏

coursera ml笔记

13 篇文章 0 订阅

订阅专栏

欢迎点击作者原文地址
所有公式速查
注意：所有的公式优先以向量的形式表示，其中

θ = (θ 0, θ 1, . . ., θ n) T

$\theta = (\theta_0,\theta_1,...,\theta_n)^T$

X = ⎡ ⎣ ⎢ ⎢ ⎢ ⎢ ⎢ \dots (x (1)) T \dots \dots (x (2)) T \dots ⋮ \dots (x (n)) T \dots ⎤ ⎦ ⎥ ⎥ ⎥ ⎥ ⎥

$X=\begin{bmatrix} \cdots (x^{(1)})^T \cdots \\ \cdots (x^{(2)})^T \cdots \\ \vdots \\ \cdots (x^{(n)})^T \cdots \end{bmatrix}$

$x^{(i)} = (x_0^{(i)},x_1^{(i)},...,x_n^{(i)})^T$
$\forall x_0^{(i)}=0$
这里的X并不是一个列向量，所以不能直接使用 $\theta^TX$ 来表示 $h_\theta(x)$

———————–	Linear Regression	Logistic Regreesion	Neural Network
Hypothesis $h θ (x)$ $h_\theta(x)$	$h θ (x) = X θ$ $h_\theta(x) = X\theta$	$h θ (x) = g (X θ) = 1 1 + e - X θ$ $h_\theta(x)=g(X\theta)=\frac{1}{1+e^{-X\theta}}$	$h θ (x)$ $h_\theta(x)$
Cost Function $J (θ)$ $J(\theta)$	$1 2 m (X θ - Y) T (X θ - Y) = 1 2 m \sum i = 1 m (h θ (x (i)) - y (i)) 2$ $\frac{1}{2m}(X\theta-Y)^T(X\theta-Y)=\frac{1}{2m}\sum_{i=1}^m(h_\theta(x^{(i)})-y^{(i)})^2$	$- 1 m [Y T l o g (X θ) + (1 - Y) T l o g (1 - X θ)] = - 1 m [\sum i = 1 m y (i) l o g h θ (x (i)) - (1 - y (i)) l o g (1 - h θ (x (i))) 2]$ $-\frac{1}{m}\left[Y^Tlog(X\theta)+(1-Y)^Tlog(1-X\theta)\right]=-\frac{1}{m}\left[\sum_{i=1}^my^{(i)}logh_\theta(x^{(i)})-(1-y^{(i)})log(1-h_\theta(x^{(i)}))^2\right]$	WAIT TO FUIFILL
Regularized Cost Function $J λ (θ)$ $J_\lambda(\theta)$	$J (θ) + λ 2 m \sum j = 1 m θ 2 j$ $J(\theta)+\frac{\lambda}{2m}\sum_{j=1}^m\theta_j^2$	$J (θ) + λ 2 m (θ T θ - θ 20)$ $J(\theta)+\frac{\lambda}{2m}(\theta^T\theta-\theta_0^2)$	WAIT TO FUIFILL
Parameters $θ j = θ j - α \partial J ( θ ) \partial θ j$ $\theta_j=\theta_j-\alpha\frac{\partial{J(\theta)}}{\partial{\theta_j}}$	$θ = θ - α m X T (X θ - Y)$ $\theta = \theta-\frac{\alpha}{m}X^T(X\theta-Y)$ ———————– $θ = (X T X) - 1 X T Y$ $\theta = (X^TX)^{-1}X^TY$	$θ = θ - α m X T (g (X θ) - Y)$ $\theta = \theta-\frac{\alpha}{m}X^T(g(X\theta)-Y)$	WAIT TO FUILFILL
Penalized Parameters $θ λ$ $\theta_\lambda$	$θ 0 : = θ 0 - α 1 m \sum i = 1 m (h θ (x (i)) - y (i)) 2 x (i) 0$ $\theta_0 := \theta_0 - \alpha \frac{1}{m}\sum_{i=1}^m(h_\theta(x^{(i)})-y^{(i)})^2x_0^{(i)}$ $θ j : = = θ j (1 - α λ m) - α 1 m \sum i = 1 m (h θ (x (i)) - y (i)) 2 x (i) (j = 1, 2, 3, . . ., n)$ $\theta_j := =\theta_j(1-\alpha\frac{\lambda}{m}) - \alpha \frac{1}{m}\sum_{i=1}^m(h_\theta(x^{(i)})-y^{(i)})^2x^{(i)}\space\space\space (j=1,2,3,...,n)$ ——————————– $θ = ⎛ ⎝ ⎜ ⎜ ⎜ ⎜ ⎜ X T X + λ ⎡ ⎣ ⎢ ⎢ ⎢ ⎢ ⎢ 01 ⋱ 1 ⎤ ⎦ ⎥ ⎥ ⎥ ⎥ ⎥ ⎞ ⎠ ⎟ ⎟ ⎟ ⎟ ⎟ - 1 X T y$ $\theta =\left( X^{T}X+\lambda\begin{bmatrix} 0\\ & 1\\ &&&\ddots\\ &&&&1 \end{bmatrix}\right)^{-1}X^{T}y$	Same as the Left Upper One( Linear Regression Gradient Descent)	WAITE TO FULFILL