吴恩达·Machine Learning || chap17 Large scale machine learning简记

17 Large scale machine learning

17-1 Learning with large datasets

Machine learning and data

Classify between confusable words. E.g., (to, two, too), (then, than)

It’s not who has the best algorithm that wins It’s who has the most data

Learning with large datasets

17-2 Stochastic gradient descent

Learning regression with gradient descent

Stochastic gradient descent

1. Randomly shuffle (reorder) training example

2. Repeat{

   ​ for $i:=1,\cdots ,m${

   ​ ${\theta }_j={\theta }_j-\alpha (h_{\theta }(x^{(i)})-y^{(i)})x_j^{(i)}$

   ​ (for every $j=0,\cdots ,n$)

   ​ }

   }

17-3 Mini-batch gradient descent

Mini-batch gradient descent
Batch gradient descent: Use all m examples in each iteration
Stochastic gradient descent: Use 1 example in each iteration
Mini-batch gradient descent: Use b examples in each iteration

example：

Say $b=10,m=1000$
Repeat{
for $i=1,11,21,31,\cdots ,991${
${\theta }_j:={\theta }_j-\alpha \frac{1}{10}{\sum }_{k=1}^{i+9}(h_{\theta }(x^{(i)})-y^{(i)})x_j^{(i)}$
（for every j=0，…,n）
}
}

17-4 Stochastic gradient descent converence

Checking for convergence

Batch gradient descent:

​ Plot $J_{train}(\theta )$ as a function of number of iterations of gradient decent.
​ $J_{train}(\theta )=\frac{1}{2m}(h_{\theta }(x^{(i)})-y^{(i)}{)}^2$

Stochastic gradient descent:

​ $cost(\theta ,(x^{(i)},y^{(i)}))=\frac{1}{2}(h_{\theta }(x^{(i)})-y^{(i)}{)}^2$

​ During learning, compute $cost(\theta ,(x^{(i)},y^{(i)}{)}^2$ before updating $\theta$ using $(x^{(i)},y^{(i)})$
​ Every 1000 iterations(say), plot $cost(\theta ,(x^{(i)},y^{(i)}{)}^2$ averaged over the last 1000 examples processed by algorithm

Learning rate a is typically held constant Can slowly decrease $\alpha$ over time if we want $\theta$ to converge. (E.g. $\alpha =\frac{const1}{iterationNumber+const2}$)

17-5 Online learning

Online learning
Shipping service website where user comes, specifies origin and destination, you offer to ship their package for some asking price, and users sometimes choose to use your shipping service (y=1) sometimes not (y=0).

Features $x$ capture properties of user, of origin/destination and asking price. We want to learn $p(y=1|x;\theta )$ to optimize price.

Other online learning example:
Product search (learning to search)
User searches for “Android phone 1080p camera”
Have 100 phones in store Will return 10 results.

$x$=features of phone, how many words in user query match name of phone, how many words in query match description of phone, etc.
$y=1$ if user clicks on link. $y=0$ otherwise
Learn $p(y=1|x;\theta )$
Use to show user the 10 phones they’re most likely to click on.

Other examples: Choosing special offers to show user; customized selection of news articles; product recommendation

17-6 Map-reduce and data parallelism

减少映射和数据并行

Map-reduce

Batch gradient descent： $${\theta }_j:={\theta }_j-\alpha \frac{1}{400}\sum _{i=1}^{400}(h_{\theta }(x^{(i)})-y^{(i)})x_j^{(i)}$$

Map-reduce and summation over the training set
Many learning algorithms can be expressed as computing sums of functions over the training set

Multi-core machines