1
00:00:00,000 --> 00:00:02,000
Hello my friends, welcome to this course.

2
00:00:02,000 --> 00:00:02,733
My name is Adil.

3
00:00:02,733 --> 00:00:05,833
On the point of this,
I am the second instructor of this course

4
00:00:05,833 --> 00:00:06,800
and it is with me

5
00:00:06,800 --> 00:00:11,000
that you will be coding all the machine
learning models of this course from parts

6
00:00:11,000 --> 00:00:12,000
one, two parts in.

7
00:00:12,000 --> 00:00:14,666
And so here
that's the beginning of our journey.

8
00:00:14,666 --> 00:00:18,466
It is this page containing
all the resources of this course,

9
00:00:18,600 --> 00:00:22,600
including the course slides
and references, the Colab notebooks

10
00:00:22,600 --> 00:00:25,600
on which we will
be coding our models in Python.

11
00:00:25,800 --> 00:00:28,166
And you also have the whole machine
learning.

12
00:00:28,166 --> 00:00:33,133
It is that folder containing all the codes
in Python, and are all the data

13
00:00:33,133 --> 00:00:37,500
sets and colorblind friendly images
whenever we do some visualization.

14
00:00:37,600 --> 00:00:38,000
All right.

15
00:00:38,000 --> 00:00:41,666
So first thing first, make sure to have
downloaded the slides on your machine

16
00:00:41,833 --> 00:00:44,833
as well as the zip folder
containing everything.

17
00:00:44,966 --> 00:00:46,133
Now it is downloaded.

18
00:00:46,133 --> 00:00:47,466
And that's the first thing
I'm going to do here.

19
00:00:47,466 --> 00:00:49,766
I'm going to introduce this folder
very quickly.

20
00:00:49,766 --> 00:00:52,600
So here it is machine learning. It is
it codes in data set.

21
00:00:52,600 --> 00:00:56,233
You have the ten parts
containing all the codes and data set.

22
00:00:56,233 --> 00:00:58,900
So for example let's go to part
three classification

23
00:00:58,900 --> 00:01:01,633
then into the logistic regression
model folder.

24
00:01:01,633 --> 00:01:02,500
And then there you go.

25
00:01:02,500 --> 00:01:06,900
You will find the codes in Python
and in R and in Python.

26
00:01:07,033 --> 00:01:10,166
Know that you have the two formats ipynb.

27
00:01:10,166 --> 00:01:14,166
If you want to code on Google Colab
or Jupyter Notebook, and py

28
00:01:14,233 --> 00:01:17,800
if you want to code on another ID
like Spyder in Anaconda.

29
00:01:17,800 --> 00:01:18,133
All right.

30
00:01:18,133 --> 00:01:19,000
And of course in each folder

31
00:01:19,000 --> 00:01:22,000
you have two data
sets that we will use to train the model.

32
00:01:22,133 --> 00:01:24,366
Okay. So now let's go back to the page.

33
00:01:24,366 --> 00:01:30,300
And now what I want to do is introduce
to you Google Colab, because it is for me

34
00:01:30,300 --> 00:01:34,800
the best interface to start with
when coding machine learning models.

35
00:01:34,800 --> 00:01:38,733
And the reason for this is that not only
it is super user friendly, you will see,

36
00:01:38,933 --> 00:01:43,166
but also the beauty is that
everything is already pre-installed

37
00:01:43,166 --> 00:01:47,866
so we won't have anything to worry about
when it comes to installing packages.

38
00:01:47,866 --> 00:01:48,600
I'll show this to you.

39
00:01:48,600 --> 00:01:52,266
In this tutorial, you will see that
TensorFlow for Deep learning is already

40
00:01:52,266 --> 00:01:55,900
installed, XGBoost is already installed,
Lightgbm is already installed.

41
00:01:55,900 --> 00:01:58,533
All the powerful
libraries are used to code.

42
00:01:58,533 --> 00:02:00,866
Machine
learning models are already pre-installed,

43
00:02:00,866 --> 00:02:03,633
so we can just import them
and directly code.

44
00:02:03,633 --> 00:02:07,466
And that's awesome compared to having to,
you know, install packages

45
00:02:07,466 --> 00:02:09,866
or libraries in the terminal. All right.

46
00:02:09,866 --> 00:02:13,000
So this is exactly what you should get
when opening

47
00:02:13,000 --> 00:02:14,466
the link of the Colab notebooks.

48
00:02:14,466 --> 00:02:17,566
To show you this,
I created a brand new Gmail address,

49
00:02:17,700 --> 00:02:21,100
rent Machine Learning at gmail.com
so that we can all start

50
00:02:21,100 --> 00:02:25,166
from the same page, and so that I can show
you how to use Google Colab.

51
00:02:25,166 --> 00:02:27,966
And to do so, let's
go to part three classification.

52
00:02:27,966 --> 00:02:32,000
I'm going to illustrate this with again
let's take the logistic regression model.

53
00:02:32,266 --> 00:02:33,800
So then you go to Python.

54
00:02:33,800 --> 00:02:38,433
And in this folder you will find two files
two logistic regression, the Ipynb

55
00:02:38,633 --> 00:02:43,333
notebook of the whole logistic regression
model and the data set okay.

56
00:02:43,333 --> 00:02:45,900
So now we're going to open the notebook
in Google Colab.

57
00:02:45,900 --> 00:02:48,833
But first before you do this
make sure to be connected

58
00:02:48,833 --> 00:02:51,600
to your Google account
with your Gmail address. Right.

59
00:02:51,600 --> 00:02:54,700
Otherwise you cannot
open the notebook in Collaboratory.

60
00:02:54,933 --> 00:02:57,900
And if you don't have a Gmail dress
with a Google account,

61
00:02:57,900 --> 00:02:59,200
then that's totally fine.

62
00:02:59,200 --> 00:03:02,200
That's exactly why we gave you
the Pi and Ipy.

63
00:03:02,200 --> 00:03:03,766
We found the whole machine learning.

64
00:03:03,766 --> 00:03:07,133
It is a zipped folder so that you can
open them in your favorite ID.

65
00:03:07,633 --> 00:03:09,600
All right. So now let's open the notebook.

66
00:03:09,600 --> 00:03:14,066
It will open the model in Collaboratory
or Google Colab.

67
00:03:14,300 --> 00:03:16,200
And now the first very important thing

68
00:03:16,200 --> 00:03:20,100
to understand is that this is a read
only notebook.

69
00:03:20,100 --> 00:03:24,300
In other words, you can't modify it,
you can't make a modification and save it.

70
00:03:24,300 --> 00:03:25,266
And the reason for this is,

71
00:03:25,266 --> 00:03:28,466
of course, that you all have access
to the same Google Drive folder.

72
00:03:28,633 --> 00:03:30,633
So of course no one can make modification.

73
00:03:30,633 --> 00:03:31,600
But no worries.

74
00:03:31,600 --> 00:03:34,800
What we'll do in the course
is that we will recode each of the machine

75
00:03:34,800 --> 00:03:37,600
learning models
so that you can really learn by doing.

76
00:03:37,600 --> 00:03:41,300
And to do so, what you have to do
is create a copy of this notebook.

77
00:03:41,300 --> 00:03:45,566
And so you go to file
and then click save a copy and drive.

78
00:03:45,566 --> 00:03:47,666
And as you can see this is creating a copy

79
00:03:47,666 --> 00:03:50,533
in which you will be able
to make some modifications.

80
00:03:50,533 --> 00:03:51,100
All right.

81
00:03:51,100 --> 00:03:52,066
So let me show you this.

82
00:03:52,066 --> 00:03:55,933
First we're going to click this little
folder here to import this data set.

83
00:03:56,100 --> 00:03:58,966
To do
so we then click this upload button here.

84
00:03:58,966 --> 00:04:01,633
Then you're going to find the machine
learning in that folder.

85
00:04:01,633 --> 00:04:04,000
And then you're going to go to part
three classification.

86
00:04:04,000 --> 00:04:06,433
Then logistic regression Python.

87
00:04:06,433 --> 00:04:09,000
And then you're going
to upload the social network

88
00:04:09,000 --> 00:04:12,000
that which is the data
set for the logistic regression model.

89
00:04:12,200 --> 00:04:16,133
Then you click open
and it will open right here the data set.

90
00:04:16,133 --> 00:04:17,100
There you go.

91
00:04:17,100 --> 00:04:19,300
And then now just for fun
let's run this whole code.

92
00:04:19,300 --> 00:04:22,033
So we click run time here and then run.

93
00:04:22,033 --> 00:04:25,533
Oh and as you can see
all the cells are now running one by one.

94
00:04:25,866 --> 00:04:30,233
And at the end well we will see
you know, all the predictions.

95
00:04:30,233 --> 00:04:31,866
And right now this cell is coding.

96
00:04:31,866 --> 00:04:33,933
It will plot the training set results

97
00:04:33,933 --> 00:04:37,700
of our logistic regression model
and the test set results just below.

98
00:04:37,900 --> 00:04:39,166
All right there we go.

99
00:04:39,166 --> 00:04:41,100
So you see that's how Google Colab works.

100
00:04:41,100 --> 00:04:44,666
Now note that this copy of this

101
00:04:44,666 --> 00:04:48,066
logistic regression
notebook is created in your drive.

102
00:04:48,066 --> 00:04:51,300
More specifically in this folder
Colab notebooks.

103
00:04:51,433 --> 00:04:52,800
If we double click on it, there you go.

104
00:04:52,800 --> 00:04:55,800
You will find the copy
that we just created.

105
00:04:56,000 --> 00:04:56,433
Okay.

106
00:04:56,433 --> 00:05:00,233
And now to finish as I promised,
I want to show you this

107
00:05:00,233 --> 00:05:03,333
beauty of Google Colab, which is the fact
that everything is already pre-installed.

108
00:05:03,333 --> 00:05:08,800
So let's open a new notebook and let's for
example, import TensorFlow or XGBoost.

109
00:05:08,933 --> 00:05:12,866
You will see that it is already here
and we don't even have to install it.

110
00:05:13,000 --> 00:05:15,900
So there we go import TensorFlow.

111
00:05:15,900 --> 00:05:17,233
That's the first one.

112
00:05:17,233 --> 00:05:18,600
Let's run the cell.

113
00:05:18,600 --> 00:05:21,400
And as you can see the cell was executed
successfully.

114
00:05:21,400 --> 00:05:22,333
And let's do the same.

115
00:05:22,333 --> 00:05:25,333
For example with XY boost XGBoost.

116
00:05:25,400 --> 00:05:28,300
Let's play the cell and same.

117
00:05:28,300 --> 00:05:32,100
It will import extra boost without
telling us that we need to install it.

118
00:05:32,100 --> 00:05:33,466
Right. There you go.

119
00:05:33,466 --> 00:05:34,833
So that's the beauty of Google Colab.

120
00:05:34,833 --> 00:05:38,300
We will directly be able to code
without worrying about anything.

121
00:05:38,433 --> 00:05:41,566
And so I look forward
to starting the implementations with you.

122
00:05:41,700 --> 00:05:43,966
And until then, enjoy machine learning.