Machine Learning on Coursera

Week Six

F Score

$$\begin{array}{rlr}P& =& \frac{2}{\frac{1}{P}+\frac{1}{R}}\\ & =& 2\frac{PR}{P+R}\end{array}$$

Week Seven

Support Vector Machine

Cost Function

$$\begin{array}{rl}& \underset{\theta }{\min }[-\frac{1}{m}\sum _{y_i\in Y,x_i\in X}{y}_i\log h({\theta }^T{x}_i)+(1-y_i)\log (1-h({\theta }^T{x}_i))+\frac{\lambda }{2m}\sum _{{\theta }_i\in \theta }{\theta }_i^2]\\ & \Rightarrow \underset{\theta }{\min }[-\sum _{y_i\in Y,x_i\in X}{y}_i\log h({\theta }^T{x}_i)+(1-y_i)\log (1-h({\theta }^T{x}_i))+\frac{\lambda }{2}\sum _{{\theta }_i\in \theta }{\theta }_i^2]\\ & \Rightarrow \underset{\theta }{\min }[C\sum _{y_i\in Y,x_i\in X}{y}_i\log h({\theta }^T{x}_i)+(1-y_i)\log (1-h({\theta }^T{x}_i))+\sum _{{\theta }_i\in \theta }{\theta }_i^2]\end{array}$$
C is somewhat $\frac{1}{\lambda }$.

• Large C:
  • lower bias, high variance
• Small C:
  • Higher bias, low variance
• Large ${\sigma }^2$: Features $f_i$ vary more smoothly.
  • Higher bias, low variance
• Small ${\sigma }^2$: Features $f_i$ vary more sharply.
  • Lower bias, high variance.
    $$\begin{array}{rlrl}& \frac{1}{2}\sum _{{\theta }_i\in \theta }{\theta }_i^2& & \\ & s.t& {\theta }^T{x}_i\ge 1,if{y}_i=1& \\ & & {\theta }^T{x}_i\le -1,if{y}_i=0& \end{array}$$

PS

If features are too many related to m, use logistic regression or SVM without a kernel.

If n is small, m is intermediate, use SVM with Gaussian kernal.

If n is small, m is large, add more features and use logistic regression or SVM without a kernel.

Week Eight

K-means

Cost Function

It try to minimize
$$\underset{\mu }{\min }\frac{1}{m}\sum _{i=1}^m||x^{(i)}-{\mu }_{c^{(i)}}|{|}^2$$
For the first loop, minimize the cost function by varing the centorid. For the second loop, it minimize the cost funcion with cetorid fixed and realign the centorid of every x in the training set.

Initialize

Initialize the centorids randomly. Randomly select k samples from the training set and set the centorids to these random selected samples.

It is possible that K-meas fall into the local minimum, So repeat to initialize the centorids randomly until the cost(distortion) is suitable for your purposes.

K-means converge all the time and it will not increase the cost during the training processs. More centoirds will decease the cost, if not, the k-means must fall into the local minimum and reinitialize the centorid until the cost is less.

PCA (Principal Component Analysis)

Restruct x from z meeting the below nonequation
$$1-\frac{\frac{1}{m}{\sum }_{i=1}^m||x^{(i)}-x_{approximation}^{(i)}|{|}^2}{\frac{1}{m}{\sum }_{i=1}^m||x^{(i)}|{|}^2}\ge 0.99$$
PS:
the nonequation can be equal to the below
$$\begin{array}{rl}[U,S,D]& =svd(sigma)\\ U_{reduce}& =U(:,1:k)\\ z& =U_{reduce}^{\prime }\ast x\\ x_{approximation}& =U_{reduce}\ast x\\ & \\ S& =\left(\begin{array}{cccc}{s}_{11}& 0& \cdots & 0\\ 0& s_{22}& \cdots & 0\\ ⋮& ⋮& \ddots & ⋮\\ 0& 0& \cdots & s_{nn}\end{array}\right)\\ & \\ \frac{{\sum }_{i=1}^k{s}_{ii}^2}{{\sum }_{i=1}^m{s}_{ii}^2}& \ge 0.99\end{array}$$

Week Nine

Anomaly Detection

Gaussian Distribution

Multivariate Gaussian Distribution takes the connection of different variants into account
$$p(x)=\frac{1}{(2\pi {)}^{\frac{n}{2}}|\Sigma {|}^{\frac{1}{2}}}{e}^{-\frac{1}{2}(x-\mu {)}^T{\Sigma }^{-1}(x-\mu )}$$
Single variant Gaussian Distribution is a special example of Multivariate Gaussian Distribution, where
$$\Sigma =\left(\begin{array}{ccccc}{\sigma }_{11}& & & & \\ & {\sigma }_{22}& & & \\ & & \ddots & & \\ & & & {\sigma }_{nn}& \end{array}\right)$$
When training the Anomaly Detection, we can use Maximum Likelihood Estimation
$$\begin{array}{rl}\mu & =\frac{1}{m}\sum _{i=1}^m{x}^{(i)}\\ \Sigma & =\frac{1}{m}\sum _{i=1}^m(x^{(i)}-\mu )(x^{(i)}-\mu {)}^T\end{array}$$
When we use single variant anomaly detection, the numerical cost is much cheaper than multivariant. But may need to add some new features to distinguish the normal and non-normal.

Recommender System

Cost Function

J ( X , Θ ) = 1 2 ∑ ( i , j ) : r ( i , j ) = 1 ( ( θ ( j ) ) T x ( i ) − y ( i , j ) ) 2 + λ 2 [ ∑ i = 1 n m ∑ k = 1 n ( x k ( i ) ) 2 + ∑ j = 1 n μ ∑ k = 1 n ( θ k ( j ) ) 2 ] J ( X , Θ ) = 1 2 S u m { ( X Θ ′ − Y ) . ∗ R } + λ 2 ( S u m { Θ . 2 } + S u m { X . 2 } \begin{aligned} J(X,\Theta) = \dfrac{1}{2} \sum_{(i,j):r(i,j)=1}((\theta^{(j)})^{T}x^{(i)}-y^{(i,j)})^2 + \dfrac{\lambda}{2}[\sum_{i=1}^{n_{m}}\sum_{k=1}^{n}(x_k^{(i)})^2 + \sum_{j=1}^{n_\mu} \sum_{k=1}^n(\theta_{k}^{(j)})^2]\\ J(X,\Theta) = \dfrac{1}{2}Sum\{(X\Theta'-Y).*R\} + \dfrac{\lambda}{2}(Sum\{\Theta.^2\} + Sum\{X.^2\}\\ \end{aligned} J(X,Θ)=21​(i,j):r(i,j)=1∑​((θ(j))Tx(i)−y(i,j))2+2λ​[i=1∑nm​​k=1∑n​(xk(i)​)2+j=1∑nμ​​k=1∑n​(θk(j)​)2]J(X,Θ)=21​Sum{(XΘ′−Y).∗R}+2λ​(Sum{Θ.2}+Sum{X.2}​
$$\begin{array}{r}\frac{\partial J}{\partial X}=((X{\Theta }^{\prime }-Y).\ast R)\Theta +\lambda X\\ \frac{\partial J}{\partial \Theta }=((X{\Theta }^{\prime }-Y).\ast R{)}^{\prime }X+\lambda \Theta \end{array}$$