1
00:00:00,750 --> 00:00:06,660
I read about an issue like, why are new roads are really cool?

2
00:00:07,050 --> 00:00:11,750
We cannot just use a single neuron to perform complex has.

3
00:00:12,390 --> 00:00:18,990
This is the reason our brain had billions of neurons stacked in layers forming a network.

4
00:00:19,010 --> 00:00:22,740
Similarly, artificial neurons are engaged in layers.

5
00:00:23,220 --> 00:00:29,690
Each and every layer will be connected in such a way that information is bought from one to another.

6
00:00:29,850 --> 00:00:38,130
The typical Annand consists of the following layers input layer upon layer and about layer.

7
00:00:38,700 --> 00:00:44,190
So Layer has a collection of neurons and are neurons in one layer interact.

8
00:00:45,190 --> 00:00:53,740
With all the new rules in other areas, however, new rules in the assembly will not interact with one

9
00:00:53,740 --> 00:01:04,090
another simply because new rules from the adjacent layers have connection between the however new rules

10
00:01:04,090 --> 00:01:04,820
in the same place.

11
00:01:04,860 --> 00:01:12,640
You do not have any connections with your thumb, nor are you to present the new rules in the artificial

12
00:01:12,640 --> 00:01:13,450
neural network.

13
00:01:13,870 --> 00:01:16,970
And a typical animation in this.

14
00:01:20,280 --> 00:01:28,080
As can you say, that in both hit and light and how put your shoulder input layer is where we fit in

15
00:01:28,270 --> 00:01:35,190
to the network, the numbers up newroz it is the number of in both we feed to the network.

16
00:01:35,580 --> 00:01:39,180
Each input will have some influence on predicting the output.

17
00:01:39,690 --> 00:01:44,270
However, no computation is performed in the input layer.

18
00:01:44,610 --> 00:01:49,020
It is just you for passive information from the outside world to the network.

19
00:01:50,240 --> 00:01:57,890
He didn't lie any lie between the input layer and the output layer is called a layer is processed.

20
00:01:57,890 --> 00:02:05,390
The input received from the input layer, the hidden layer is responsible for the arriving complex relationships

21
00:02:05,870 --> 00:02:07,600
between input and output.

22
00:02:08,090 --> 00:02:10,250
That is, the hidden layer identified.

23
00:02:10,250 --> 00:02:16,190
A pattern in the data set is a majorly responsible for learning the data representation.

24
00:02:17,080 --> 00:02:18,970
And for extracting the features.

25
00:02:20,240 --> 00:02:26,780
There can be any number of years, however, we have to to a number of Hidalgo's, according to our

26
00:02:26,870 --> 00:02:27,800
use case.

27
00:02:28,730 --> 00:02:35,990
For a very simple problem, we can just use one dollar, but while performing complex task such as image

28
00:02:35,990 --> 00:02:36,800
recognition.

29
00:02:37,680 --> 00:02:46,140
With so many hits like yours where failure is responsible for extracting important features, the network

30
00:02:46,710 --> 00:02:51,900
is Konar deep neural network when we have many hit delayers.

31
00:02:52,820 --> 00:03:00,590
About layer after processing the input, the hidden layers and its result to the output layer and the

32
00:03:00,590 --> 00:03:07,130
name suggests I like it missed the output, the number of neurons in the output layer is based on the

33
00:03:07,130 --> 00:03:08,100
tile problem.

34
00:03:08,480 --> 00:03:11,140
We want our network to solve it.

35
00:03:11,450 --> 00:03:19,040
If it's a binary classification, then a number of new roles in the Abulafia is one that tell us which

36
00:03:19,040 --> 00:03:20,810
class in Bush belong.

37
00:03:21,820 --> 00:03:31,420
It is a Montebourg multiclass classification, say, with five glasses, and we want to get the probability

38
00:03:31,420 --> 00:03:39,640
of a glass as an output than the number of new in the output layer is five ish, emitting the probability

39
00:03:40,240 --> 00:03:49,060
it is a regression problem, then we have one neuron in the Abulafia and that is on about and I hope

40
00:03:49,060 --> 00:03:52,450
you enjoy it and I will see you in the next video.