1
00:00:11,110 --> 00:00:14,890
In this video, we will be introduced to the next section of this course.

2
00:00:15,670 --> 00:00:18,880
The main topic of this section is recurrent neural networks.

3
00:00:19,420 --> 00:00:21,070
We also call these aren't ends.

4
00:00:21,850 --> 00:00:23,440
So what makes Arnold special?

5
00:00:24,250 --> 00:00:26,860
Well, consider the neural networks we have studied so far.

6
00:00:27,310 --> 00:00:32,590
They all have this one peculiar property, which makes them unlike biological neural networks.

7
00:00:33,250 --> 00:00:38,320
And this is that all the information goes in the same direction from left to right.

8
00:00:39,130 --> 00:00:44,800
So here's a question to consider why not simply have neuronal connections which are less restrictive,

9
00:00:45,190 --> 00:00:50,740
for example, ones that can go backwards when we create connections that go backwards in a specific

10
00:00:50,740 --> 00:00:52,900
way with a time delay of one step?

11
00:00:53,170 --> 00:00:55,330
We call these recurrent neural networks.

12
00:00:56,230 --> 00:01:01,090
Now, please note that there are all kinds of ways we could make new connections, but only this specific

13
00:01:01,090 --> 00:01:03,520
way is what we call a recurrent neural network.

14
00:01:04,569 --> 00:01:10,090
OK, so this diagram is pretty messy, but what you should notice is that every hidden unit has an arrow

15
00:01:10,210 --> 00:01:13,760
that goes back towards all other hidden units, including itself.

16
00:01:14,620 --> 00:01:20,200
The most common diagram, which looks like this is a bit cleaner but actually does not reflect how the

17
00:01:20,200 --> 00:01:22,660
connections are made in this case.

18
00:01:22,690 --> 00:01:25,990
Notice how each heading unit loops back only to itself.

19
00:01:26,380 --> 00:01:28,150
This is not exactly what we want.

20
00:01:33,050 --> 00:01:36,930
So let's do a brief outline of this section as per usual.

21
00:01:37,010 --> 00:01:40,970
This section will start with the theory and concepts behind Arnett's.

22
00:01:41,390 --> 00:01:47,360
Well, first study the simple Arnon, also known as the eminent in this will form the basis for everything

23
00:01:47,360 --> 00:01:48,050
else we do.

24
00:01:49,310 --> 00:01:54,770
Once we understand the simple answer and we'll do a small code exercise which will reinforce your understanding.

25
00:01:55,880 --> 00:02:00,320
As you recall, one of the major themes of this course is thinking about shapes.

26
00:02:00,830 --> 00:02:03,830
This section will really test your ability to understand this.

27
00:02:04,610 --> 00:02:09,050
Now, when it comes to art ends, the simple answer is rarely ever used any more.

28
00:02:09,949 --> 00:02:16,550
For the time being, the LACMA is the most common type of art in unit, so we'll study the LACMA and

29
00:02:16,550 --> 00:02:18,560
the closely related GIU.

30
00:02:19,040 --> 00:02:26,480
You'll see why the LACMA and Giu are needed and what advantages they have over the simple art in the

31
00:02:26,480 --> 00:02:31,220
next step will be to put everything into practice with various data sets as usual.

32
00:02:32,120 --> 00:02:36,320
Now I want you to remember one thing, which is that all data is the same.

33
00:02:36,830 --> 00:02:42,050
So although earnings can be applied to all of the data sets we've looked at so far, we of course do

34
00:02:42,050 --> 00:02:43,340
not have time for the mall.

35
00:02:43,790 --> 00:02:48,860
So if there's anything you want to try them on, please remember that no code has to change.

36
00:02:49,190 --> 00:02:51,710
Simply switch one data set out for another.

37
00:02:52,220 --> 00:02:56,690
And of course, there will be plenty of other configuration options to choose from as well.

38
00:02:57,080 --> 00:02:59,750
So please feel free to fiddle around and test things out.

39
00:03:04,560 --> 00:03:09,510
Now, I want to make a special note that like with convolution, you don't have to go through the theory

40
00:03:09,510 --> 00:03:12,600
lectures if you find the math is intimidating for you.

41
00:03:13,050 --> 00:03:18,150
Instead, if you prefer to treat the code simply like an API, you can do that as well.

42
00:03:18,660 --> 00:03:22,740
So if that's the strategy you wish to take, please skip to the code preparation.

43
00:03:24,210 --> 00:03:29,670
In my opinion, almonds contain a bit less math and science, so they might be easier to understand.

44
00:03:30,270 --> 00:03:33,810
It's mostly about understanding the diagrams and the flow of the data.

45
00:03:34,530 --> 00:03:38,370
However, understanding the equations you see will make it much simpler.

46
00:03:39,270 --> 00:03:44,490
This actually goes back to our discussion on machine learning, where I said that one of the most important

47
00:03:44,490 --> 00:03:49,790
expressions you will learn is W Transpose X plus B for Arnaz.

48
00:03:49,800 --> 00:03:52,320
We are going to see this in multiple places at once.

49
00:03:52,890 --> 00:03:58,290
Recall that I advised you not to think about this as math, but merely a pattern or an operation.

50
00:03:58,950 --> 00:04:04,200
So once you see things from this perspective, it makes it easier since you can think at a higher level

51
00:04:04,350 --> 00:04:09,420
about how these more fundamental operations can be connected together to make new things.

52
00:04:10,110 --> 00:04:15,360
In fact, I encourage you to watch this section with the perspective that there actually is no math

53
00:04:15,360 --> 00:04:19,440
at all, although you're going to see some equations that look like math.

54
00:04:19,649 --> 00:04:23,820
It's actually very interesting to note that you don't have to actually do any math.

55
00:04:24,390 --> 00:04:28,770
The section is more about understanding things at a higher level and how different building blocks fit

56
00:04:28,770 --> 00:04:31,050
together to create bigger systems.

57
00:04:31,650 --> 00:04:36,270
Of course, it's possible you might be so intimidated by the math that you'll see it in.

58
00:04:36,450 --> 00:04:40,830
All, this advice will go out the window so you can make that determination for yourself.