﻿1
00:00:00,166 --> 00:00:02,700
Hello and welcome to this art tutorial.

2
00:00:02,700 --> 00:00:06,566
So in a previous tutorial,
we just pre-processed our data set to make

3
00:00:06,566 --> 00:00:10,833
it ready for our logistic regression model
that we will fit on the training set.

4
00:00:11,266 --> 00:00:13,366
So that's what
we're going to do in this tutorial.

5
00:00:13,366 --> 00:00:14,866
And let's do it right now.

6
00:00:14,866 --> 00:00:16,333
So it's going to be very simple.

7
00:00:16,333 --> 00:00:18,933
We're just going to create our classifier

8
00:00:20,466 --> 00:00:21,466
equals.

9
00:00:21,466 --> 00:00:25,766
And then we're going to use an existing
function of R called the glm function.

10
00:00:26,000 --> 00:00:29,100
That is going to build this
logistic regression classifier.

11
00:00:29,666 --> 00:00:31,800
So let's type here glm.

12
00:00:31,800 --> 00:00:35,700
And then let's press F1
to see what parameters we have to input.

13
00:00:36,433 --> 00:00:36,733
Okay.

14
00:00:36,733 --> 00:00:41,233
So the first thing we see is
that GLM is for generalized linear models.

15
00:00:41,233 --> 00:00:45,000
And that's because the logistic regression
is a linear classifier.

16
00:00:45,100 --> 00:00:47,466
For those of you
who didn't follow the Python tutorial,

17
00:00:47,466 --> 00:00:50,133
you will see that our logistic regression
classifier

18
00:00:50,133 --> 00:00:53,700
will linearly separate
our two classes of users.

19
00:00:54,366 --> 00:00:57,366
And then let's look at the arguments
to see why we have two inputs.

20
00:00:58,800 --> 00:01:01,233
All right
so the first argument is formula.

21
00:01:01,233 --> 00:01:03,900
So that's the same as what we did
in the regression section here.

22
00:01:03,900 --> 00:01:05,966
We're going to write formula

23
00:01:07,166 --> 00:01:08,266
equals.

24
00:01:08,266 --> 00:01:11,933
And then we take the dependent variable
which is purchased.

25
00:01:13,533 --> 00:01:15,600
Then alt plus n.

26
00:01:15,600 --> 00:01:18,900
And then here a dot to specify
that we want to take all

27
00:01:18,900 --> 00:01:20,600
the independent variables.

28
00:01:20,600 --> 00:01:23,900
So that's way
R understands that we want to predict

29
00:01:24,100 --> 00:01:28,033
the purchased result
based on all the independent variables

30
00:01:28,033 --> 00:01:31,633
which are the age
and the estimated salary.

31
00:01:33,000 --> 00:01:33,533
All right.

32
00:01:33,533 --> 00:01:35,333
Now what is the next argument.

33
00:01:35,333 --> 00:01:37,966
So let's add a comma here. Enter.

34
00:01:37,966 --> 00:01:40,066
And let's add the second argument.

35
00:01:40,066 --> 00:01:41,900
So the second argument is family.

36
00:01:41,900 --> 00:01:46,400
And here since we are in logistic
regression we're going to add family

37
00:01:47,500 --> 00:01:50,500
equals binomial.

38
00:01:51,700 --> 00:01:52,900
So at this stage it's

39
00:01:52,900 --> 00:01:56,066
not important to understand
what binomial means exactly.

40
00:01:56,433 --> 00:01:57,800
But you just have to remember that

41
00:01:57,800 --> 00:02:01,666
for logistic regression
you have to specify the binomial family.

42
00:02:01,666 --> 00:02:01,900
Here.

43
00:02:02,900 --> 00:02:05,966
And then
we just need to input one last argument.

44
00:02:06,166 --> 00:02:07,600
It's the data here

45
00:02:07,600 --> 00:02:11,100
which is the data on which you want to
train your logistic regression model.

46
00:02:11,500 --> 00:02:13,800
So of course that's our training set.

47
00:02:13,800 --> 00:02:17,900
So we're going to add here
data equals training set.

48
00:02:17,933 --> 00:02:20,933
Here we go. Here it is. Okay.

49
00:02:21,533 --> 00:02:22,266
And that's it.

50
00:02:22,266 --> 00:02:27,133
By only using this GLM function
and precisely the formula the family

51
00:02:27,133 --> 00:02:30,133
and the fact that we're building the model
and the training set

52
00:02:30,366 --> 00:02:33,333
that build
our logistic regression classifier.

53
00:02:33,333 --> 00:02:36,033
So now let's select this

54
00:02:36,033 --> 00:02:39,033
command and control plus enter to execute.

55
00:02:39,566 --> 00:02:40,433
And here we go.

56
00:02:40,433 --> 00:02:42,900
The classifier is built.

57
00:02:42,900 --> 00:02:43,666
Okay so that's it.

58
00:02:43,666 --> 00:02:47,000
For this tutorial we fitted logistic
regression to the training set.

59
00:02:47,100 --> 00:02:48,366
And in the next tutorial

60
00:02:48,366 --> 00:02:52,633
we will be predicting the test results
using this classifier that we just built.

61
00:02:53,200 --> 00:02:55,000
All right
thank you for watching this tutorial.

62
00:02:55,000 --> 00:02:56,800
I look forward
to seeing you in the next one.

63
00:02:56,800 --> 00:02:58,700
And until then, enjoy machine learning.
