13 - 1 - Unsupervised Learninguction (3 min).srt

1
00:00:00,090 --> 00:00:02,320
In this video, I'd like to start to talk about clustering.
在这个视频中 我将开始介绍聚类算法
(字幕整理：中国海洋大学 黄海广,haiguang2000@qq.com )

2
00:00:03,420 --> 00:00:04,850
This will be exciting because this
这将是一个激动人心的时刻

3
00:00:04,930 --> 00:00:06,910
is our first unsupervised learning algorithm
因为这是我们学习的第一个非监督学习算法

4
00:00:07,350 --> 00:00:09,080
where we learn from unlabeled data
我们将要让计算机学习无标签数据

5
00:00:09,840 --> 00:00:10,730
instead of the label data.
而不是此前的标签数据

6
00:00:11,900 --> 00:00:13,300
So, what is unsupervised learning?
那么 什么是非监督学习呢

7
00:00:14,390 --> 00:00:15,630
I briefly talked about unsupervised
在课程的一开始

8
00:00:16,350 --> 00:00:17,470
learning at the beginning of
我曾简单的介绍过非监督学习

9
00:00:17,550 --> 00:00:18,560
the class but it's useful
然而 我们还是有必要

10
00:00:19,030 --> 00:00:20,320
to contrast it with supervised learning. So, here's
将其与监督学习做一下比较

11
00:00:21,760 --> 00:00:23,750
a typical supervisory problem where
在一个典型的监督学习中

12
00:00:24,030 --> 00:00:25,470
we are given a label training sets
我们有一个有标签的训练集

13
00:00:25,770 --> 00:00:27,470
and the goal is to find
我们的目标是找到

14
00:00:27,980 --> 00:00:29,420
the decision boundary that separates the
能够区分正样本和负样本的

15
00:00:29,530 --> 00:00:31,310
positive label examples and the negative label examples.
决策边界

16
00:00:33,100 --> 00:00:34,400
The supervised learning problem in
在这里的监督学习中

17
00:00:34,460 --> 00:00:35,710
this case is given a
我们有一系列标签

18
00:00:35,850 --> 00:00:38,270
set of labels to fit a hypothesis to it.
我们需要据此拟合一个假设函数

19
00:00:39,160 --> 00:00:40,560
In contrast, in the unsupervised
与此不同的是

20
00:00:41,080 --> 00:00:42,420
learning problem, we're given
在非监督学习中

21
00:00:42,710 --> 00:00:43,740
data that does not have
我们的数据没有

22
00:00:43,890 --> 00:00:45,270
any labels associated with it.
附带任何标签

23
00:00:46,730 --> 00:00:47,940
So we're given data that looks
我们拿到的数据就是这样的

24
00:00:48,100 --> 00:00:49,090
like this, here's a set
在这里

25
00:00:49,180 --> 00:00:50,470
of points and then no labels.
我们有一系列点 却没有标签

26
00:00:51,800 --> 00:00:52,860
And so our training set is written
因此 我们的训练集可以写成

27
00:00:53,220 --> 00:00:54,720
just x1, x2 and
只有x(1) x(2)

28
00:00:55,210 --> 00:00:56,890
so on up to x(m)
一直到 x(m)

29
00:00:57,450 --> 00:00:58,720
and we don't get any labels y.
我们没有任何标签y

30
00:00:59,540 --> 00:01:00,800
And that's why the points plotted
因此 图上画的这些点

31
00:01:01,160 --> 00:01:02,300
up on the figure don't have
没有

32
00:01:02,430 --> 00:01:04,330
any labels on them.
标签信息

33
00:01:04,490 --> 00:01:05,510
So in unsupervised learning, what
也就是说 在非监督学习中

34
00:01:05,710 --> 00:01:06,860
we do is, we give this sort of
我们需要将一系列

35
00:01:07,280 --> 00:01:09,150
unlabeled training set to
无标签的训练数据

36
00:01:09,250 --> 00:01:10,510
an algorithm and we just
输入到一个算法中

37
00:01:10,600 --> 00:01:12,220
ask the algorithm: find some
然后我们告诉这个算法

38
00:01:12,430 --> 00:01:14,130
structure in the data for us.
快去为我们找找这个数据的内在结构

39
00:01:15,420 --> 00:01:16,490
Given this data set, one
给定数据

40
00:01:16,650 --> 00:01:17,810
type of structure we might
我们可能需要某种算法

41
00:01:18,010 --> 00:01:19,540
have an algorithm find, is that
帮助我们寻找一种结构

42
00:01:19,810 --> 00:01:21,440
it looks like this data set has
图上的数据看起来

43
00:01:21,620 --> 00:01:23,740
points grouped into two
可以分成两个

44
00:01:24,030 --> 00:01:25,500
separate clusters and so
分开的点集（称为簇）

45
00:01:25,800 --> 00:01:28,230
an algorithm that finds that
一个能够找到

46
00:01:28,360 --> 00:01:29,230
clusters like the ones I just
我圈出的这些点集的算法

47
00:01:29,450 --> 00:01:30,610
circled, is called a clustering
就被称为

48
00:01:32,440 --> 00:01:32,440
algorithm.
聚类算法

49
00:01:33,160 --> 00:01:34,620
And this will be our first type of
这将是我们介绍的

50
00:01:34,720 --> 00:01:36,590
unsupervised learning, although there
第一个非监督学习算法

51
00:01:36,790 --> 00:01:38,320
will be other types of unsupervised
当然 此后我们还将提到

52
00:01:39,020 --> 00:01:40,200
learning algorithms that we'll talk
其他类型的非监督学习算法

53
00:01:40,320 --> 00:01:41,710
about later that finds other
它们可以为我们

54
00:01:42,130 --> 00:01:43,710
types of structure or other
找到其他类型的结构

55
00:01:43,920 --> 00:01:46,000
types of patterns in the data other than clusters.
或者其他的一些模式 而不只是簇

56
00:01:46,900 --> 00:01:48,360
We'll talk about this afterwards, we will talk about clustering.
我们将先介绍聚类算法 此后 我们将陆续介绍其他算法

57
00:01:50,020 --> 00:01:51,210
So what is clustering good for?
好啦 那么聚类算法一般用来做什么呢？

58
00:01:51,380 --> 00:01:54,350
Early in this class I had already mentioned a few applications.
在这门课程的早些时候 我曾经列举过一些应用

59
00:01:54,950 --> 00:01:56,540
One is market segmentation, where
比如市场分割

60
00:01:56,670 --> 00:01:57,690
you may have a database of
也许你在数据库中

61
00:01:57,770 --> 00:01:58,840
customers and want to group
存储了许多客户的信息

62
00:01:59,070 --> 00:02:00,380
them into different market segments.
而你希望将他们分成不同的客户群

63
00:02:00,950 --> 00:02:02,590
So, you can sell to
这样你可以对不同类型的客户分别销售产品

64
00:02:02,720 --> 00:02:05,570
them separately or serve your different market segments better.
或者分别提供更适合的服务

65
00:02:06,730 --> 00:02:08,370
Social network analysis, there are
社交网络分析

66
00:02:08,580 --> 00:02:10,090
actually, you know, groups that have done
事实上 有许多研究人员

67
00:02:10,320 --> 00:02:12,590
this, things like looking at a
正在研究这样一些内容

68
00:02:12,730 --> 00:02:14,540
group of people, social networks,
他们关注一群人 关注社交网络

69
00:02:15,070 --> 00:02:16,390
so things like Facebook, Google plus
例如 Facebook Google+

70
00:02:16,710 --> 00:02:18,260
or maybe information about who
或者是其他的一些信息

71
00:02:18,430 --> 00:02:19,710
are the people that you
比如说

72
00:02:20,030 --> 00:02:21,110
email the most frequently and who
你经常跟哪些人联系

73
00:02:21,230 --> 00:02:22,170
are the people that they email
而这些人

74
00:02:22,310 --> 00:02:23,600
the most frequently, and
又经常给哪些人发邮件

75
00:02:23,750 --> 00:02:25,400
to find coherent groups of people.
由此找到关系密切的人群

76
00:02:26,500 --> 00:02:27,600
So, this would be another maybe
因此 这可能需要另一个

77
00:02:28,180 --> 00:02:28,850
clustering algorithm where, you know, you'd want
聚类算法

78
00:02:29,080 --> 00:02:32,200
to find who other coherent groups of friends in a social network.
你希望用它发现社交网络中关系密切的朋友

79
00:02:33,140 --> 00:02:33,990
Here's something that one of my
我有一个朋友

80
00:02:34,140 --> 00:02:35,170
friends actually worked on, which is,
正在研究这个问题

81
00:02:35,920 --> 00:02:37,200
use clustering to organize compute
他希望使用聚类算法

82
00:02:37,670 --> 00:02:39,220
clusters or to organize data
来更好的组织计算机集群

83
00:02:39,440 --> 00:02:40,600
centers better because, if you
或者更好的管理数据中心

84
00:02:40,800 --> 00:02:42,450
know which computers in the
因为如果你知道数据中心中

85
00:02:42,520 --> 00:02:44,990
data center are in the cluster there tend to work together.
那些计算机经常协作工作

86
00:02:45,400 --> 00:02:46,270
You can use that to reorganize
那么 你可以

87
00:02:46,950 --> 00:02:48,390
your resources and how you
重新分配资源

88
00:02:48,570 --> 00:02:50,120
lay out its network and
重新布局网络

89
00:02:50,260 --> 00:02:52,040
how design your data center and communications.
由此优化数据中心 优化数据通信

90
00:02:53,140 --> 00:02:54,540
And lastly something that, actually
最后 我实际上

91
00:02:54,850 --> 00:02:55,910
another thing I worked on, using
还在研究

92
00:02:56,130 --> 00:02:57,810
clustering algorithms to understand
如何利用聚类算法

93
00:02:58,400 --> 00:03:00,030
galaxy formation and using
了解星系的形成

94
00:03:00,280 --> 00:03:02,260
that to understand how, to
然后用这个知识

95
00:03:02,600 --> 00:03:03,860
understand astronomical detail.
了解一些天文学上的细节问题

96
00:03:06,550 --> 00:03:08,580
So, that's clustering which
好的 这就是聚类算法

97
00:03:08,890 --> 00:03:10,450
is our first example of
这将是我们介绍的第一个

98
00:03:10,530 --> 00:03:12,650
an unsupervised learning algorithm.
非监督学习算法

99
00:03:13,090 --> 00:03:14,200
In the next video, we'll start to
在下一个视频中 我们将开始介绍

100
00:03:14,370 --> 00:03:16,250
talk about a specific clustering algorithm.
一个具体的聚类算法