8 - 7 - Multiclass Classification (4 min).srt

1
00:00:00,320 --> 00:00:01,410
In this video, I want to
在这个视频中，我想
(字幕整理：中国海洋大学 黄海广,haiguang2000@qq.com )

2
00:00:01,490 --> 00:00:02,710
tell you about how to use neural
告诉你如何使用神经

3
00:00:02,900 --> 00:00:04,390
networks to do multiclass
网络做多类

4
00:00:04,830 --> 00:00:06,690
classification where we may
分类，我们可以

5
00:00:06,820 --> 00:00:07,840
have more than one category
有一个以上的类别

6
00:00:07,930 --> 00:00:09,600
that we're trying to distinguish amongst.
我们正在试图区分之间。

7
00:00:10,470 --> 00:00:12,280
In the last part of
中的最后部分

8
00:00:12,600 --> 00:00:13,920
the last video, where we
最后一个视频，在这里我们

9
00:00:14,400 --> 00:00:15,320
had the handwritten digit recognition
有手写数字识别

10
00:00:15,830 --> 00:00:17,030
problem, that was actually
的问题，这实际上是

11
00:00:17,700 --> 00:00:19,000
a multiclass classification problem because
因为多类分类问题

12
00:00:19,440 --> 00:00:20,730
there were ten possible categories
有十个可能的类别

13
00:00:21,550 --> 00:00:22,820
for recognizing the digits from
从认识数字

14
00:00:23,040 --> 00:00:23,980
0 through 9 and so, if
0至9，所以，如果

15
00:00:24,060 --> 00:00:25,430
you want us to fill you
您希望我们填补你

16
00:00:25,830 --> 00:00:27,840
in on the details of how to do that.
在关于如何做到这一点的细节。

17
00:00:30,410 --> 00:00:31,870
The way we do multiclass classification
我们做的多类分类的方法

18
00:00:32,990 --> 00:00:34,380
in a neural network is essentially
在一个神经网络本质上是

19
00:00:35,060 --> 00:00:37,600
an extension of the one versus all method.
一个与所有方法的延伸。

20
00:00:38,610 --> 00:00:39,650
So, let's say that we
所以，让我们说，我们

21
00:00:39,790 --> 00:00:41,660
have a computer vision example,
有计算机视觉的例子，

22
00:00:42,630 --> 00:00:43,810
where instead of just trying
在那里，而不是只是想

23
00:00:44,010 --> 00:00:46,170
to recognize cars as in
认识到汽车在

24
00:00:46,310 --> 00:00:47,290
the original example that I started off
那我开始了最初的例子

25
00:00:47,470 --> 00:00:48,670
with, but let's say that
用，但让我们说，

26
00:00:49,060 --> 00:00:51,380
we're trying to recognize, you know, four
我们试图认识，你知道，四

27
00:00:51,510 --> 00:00:52,820
categories of objects and given
对象的类别和特定

28
00:00:53,030 --> 00:00:53,900
an image we want to
我们想要的图像

29
00:00:54,100 --> 00:00:56,360
decide if it is a pedestrian, a car, a motorcycle or a truck.
决定是否有行人，汽车，摩托车或卡车。

30
00:00:57,200 --> 00:00:58,750
If that's the case, what
如果是这样的话，有什么

31
00:00:58,920 --> 00:01:00,480
we would do is we would
我们会做的是，我们会

32
00:01:00,970 --> 00:01:02,820
build a neural network with four
建立一个神经网络与四

33
00:01:03,160 --> 00:01:04,500
output units so that
输出单元，使得

34
00:01:04,710 --> 00:01:08,110
our neural network now outputs a vector of four numbers.
我们的神经网络现在输出四个数字的向量。

35
00:01:09,110 --> 00:01:10,450
So, the output now is actually
所以，现在输出实际上是

36
00:01:11,170 --> 00:01:11,840
needing to be a vector of four
无需是四的矢量

37
00:01:12,070 --> 00:01:13,300
numbers and what we're
数字和我们在做什么

38
00:01:13,540 --> 00:01:14,400
going to try to do is
要尝试做的是

39
00:01:14,780 --> 00:01:16,680
get the first output unit
获得第一输出单元

40
00:01:17,180 --> 00:01:18,840
to classify: is the
分类：是

41
00:01:19,160 --> 00:01:20,650
image a pedestrian, yes or no.
图像中的行人，yes或no。

42
00:01:21,200 --> 00:01:24,530
The second unit to classify: is the image a car, yes or no.
第二个单元进行分类：是图像一辆车，是或否。

43
00:01:25,110 --> 00:01:26,880
This unit to classify: is the
本单元进行分类：是

44
00:01:27,130 --> 00:01:29,150
image a motorcycle, yes or
像摩托车，或者是

45
00:01:29,230 --> 00:01:30,460
no, and this would classify:
没有，而这将分类：

46
00:01:30,930 --> 00:01:32,930
is the image a truck, yes or no.
是图像卡车，yes或no。

47
00:01:33,720 --> 00:01:35,730
And thus, when the image
因而，当图像

48
00:01:36,390 --> 00:01:37,630
is of a pedestrian, we
是一行人，我们

49
00:01:37,820 --> 00:01:38,930
would ideally want the network
在理想情况下希望网络

50
00:01:39,410 --> 00:01:40,140
to output 1, 0, 0, 0,
到输出1，0，0，0，

51
00:01:40,250 --> 00:01:41,260
when it is a
当它是一个

52
00:01:41,520 --> 00:01:42,310
car we want it to output
车上我们希望它输出

53
00:01:42,750 --> 00:01:43,530
0, 1, 0, 0, when this
0，1，0，0，当该

54
00:01:43,840 --> 00:01:45,960
is a motorcycle, we get it to or rather, we want
被一辆摩托车，我们得到它或者说，我们要

55
00:01:46,390 --> 00:01:47,460
it to output 0, 0,
它输出到0，0，

56
00:01:47,580 --> 00:01:48,970
1, 0 and so on.
1,0等。

57
00:01:50,750 --> 00:01:51,880
So this is just like
因此，这就像

58
00:01:52,270 --> 00:01:53,690
the "one versus all" method
“一对所有”的方法

59
00:01:54,190 --> 00:01:55,520
that we talked about when we
我们谈到我们时

60
00:01:55,680 --> 00:01:58,120
were describing logistic regression, and
被描述logistic回归，和

61
00:01:58,320 --> 00:02:00,480
here we have essentially four logistic
在这里，我们有四个基本的逻辑

62
00:02:01,290 --> 00:02:03,100
regression classifiers, each of
回归分类器，每个

63
00:02:03,260 --> 00:02:04,800
which is trying to recognize one
这是试图承认一个

64
00:02:05,000 --> 00:02:06,780
of the four classes that
四班的那

65
00:02:06,940 --> 00:02:08,830
we want to distinguish amongst.
我们要区分之间。

66
00:02:09,540 --> 00:02:10,780
So, rearranging the slide of
因此，重新排列幻灯片

67
00:02:10,860 --> 00:02:12,130
it, here's our neural network
它，这里是我们的神经网络

68
00:02:12,540 --> 00:02:14,070
with four output units and those
有四个输出单元和那些

69
00:02:14,330 --> 00:02:15,510
are what we want h
是我们想要的?

70
00:02:15,670 --> 00:02:16,790
of x to be when we
x的是，当我们

71
00:02:16,990 --> 00:02:18,930
have the different images, and
有不同的图像，并

72
00:02:19,580 --> 00:02:20,860
the way we're going to represent the
我们要代表的方式

73
00:02:21,110 --> 00:02:22,690
training set in these settings
培训在这些设置中设置

74
00:02:23,260 --> 00:02:24,670
is as follows. So, when we have
如下。所以，当我们有

75
00:02:24,890 --> 00:02:26,170
a training set with different images
训练集与不同的图像

76
00:02:27,350 --> 00:02:28,990
of pedestrians, cars, motorcycles and
的行人，汽车，摩托车及

77
00:02:29,260 --> 00:02:30,450
trucks, what we're going
卡车，我们要去

78
00:02:30,510 --> 00:02:31,940
to do in this example is
做在这个例子中是

79
00:02:32,190 --> 00:02:34,580
that whereas previously we had
而那之前我们有

80
00:02:34,990 --> 00:02:36,780
written out the labels as
写出来的标签作为

81
00:02:37,040 --> 00:02:38,320
y being an integer from
y是整数

82
00:02:38,710 --> 00:02:42,180
1, 2, 3 or 4. Instead of
1，2，3或4。相反的

83
00:02:42,280 --> 00:02:44,210
representing y this way,
代表?这样一来，

84
00:02:44,890 --> 00:02:46,340
we're going to instead represent y
我们要代替代表?

85
00:02:47,050 --> 00:02:49,400
as follows: namely Yi
如下：即易

86
00:02:54,850 --> 00:02:55,230
will be either 1, 0, 0, 0
将是1，0，0，0

87
00:02:55,230 --> 00:02:57,040
or 0, 1, 0, 0 or 0, 0, 1, 0 or 0, 0, 0, 1 depending on what the
或0，1，0，0或0，0，1，0或0，0，0，1，取决于什么样的

88
00:02:57,490 --> 00:02:59,100
corresponding image Xi is.
相应的图像Xi为。

89
00:02:59,410 --> 00:03:00,700
And so one training example
等一训练实例

90
00:03:01,230 --> 00:03:03,090
will be one pair Xi colon Yi
将一对结肠熙怡

91
00:03:04,530 --> 00:03:06,340
where Xi is an image with, you
其中Xi是一个形象，你

92
00:03:06,440 --> 00:03:08,000
know one of the four objects and
知道有四个对象，并

93
00:03:08,170 --> 00:03:09,640
Yi will be one of these vectors.
易将这些载体之一。

94
00:03:10,970 --> 00:03:12,020
And hopefully, we can find
并希望，我们可以找到

95
00:03:12,420 --> 00:03:13,670
a way to get our
一种方式来获得我们

96
00:03:14,020 --> 00:03:15,100
Neural Networks to output some
神经网络的输出部分

97
00:03:15,290 --> 00:03:16,480
value. So, the h of x
值。因此，x的?

98
00:03:17,310 --> 00:03:20,360
is approximately y and
大约是y和

99
00:03:20,550 --> 00:03:22,000
both h of x and Yi,
x和毅小时，

100
00:03:22,600 --> 00:03:23,770
both of these are going
这两个要

101
00:03:24,020 --> 00:03:25,170
to be in our example,
要在我们的例子中，

102
00:03:26,060 --> 00:03:28,700
four dimensional vectors when we have four classes.
当我们有四个类四维向量。

103
00:03:31,810 --> 00:03:33,020
So, that's how you
所以，这是你如何

104
00:03:33,170 --> 00:03:34,830
get neural network to do multiclass classification.
得到的神经网络做多类分类。

105
00:03:36,290 --> 00:03:37,780
This wraps up our discussion on
这个部分是我们的讨论

106
00:03:38,050 --> 00:03:39,620
how to represent Neural Networks
如何表示神经网络

107
00:03:40,120 --> 00:03:41,620
that is on our hypotheses representation.
这是对我们的假设表示。

108
00:03:42,780 --> 00:03:44,180
In the next set of videos, let's
在下一组的视频，让我们

109
00:03:44,690 --> 00:03:45,830
start to talk about how take
开始谈论如何利用

110
00:03:45,990 --> 00:03:47,360
a training set and how to
训练集，以及如何

111
00:03:47,570 --> 00:03:49,970
automatically learn the parameters of the neural network.
自动学习神经网络的参数。