﻿1
00:00:00,366 --> 00:00:03,066
Hello and welcome to this new section
on classification.

2
00:00:03,066 --> 00:00:04,200
Very exciting.

3
00:00:04,200 --> 00:00:08,866
Classification is a very popular
and important tool in machine learning.

4
00:00:08,866 --> 00:00:10,866
So I'm super excited to get started.

5
00:00:10,866 --> 00:00:12,466
Let's have a look.

6
00:00:12,466 --> 00:00:12,833
All right.

7
00:00:12,833 --> 00:00:17,200
So classification can be defined
as a machine learning technique to

8
00:00:17,200 --> 00:00:21,833
identify the category of new observations
based on training data.

9
00:00:22,433 --> 00:00:25,766
this is different to regression where
we had to predict a continuous number.

10
00:00:25,966 --> 00:00:29,100
Here we use classification
to predict a category.

11
00:00:29,500 --> 00:00:32,466
Another important thing is that it's
a type of supervised learning algorithm.

12
00:00:32,466 --> 00:00:36,600
We'll discuss this a bit more when we
reach the next section on clustering.

13
00:00:37,000 --> 00:00:41,800
And there are actually a variety
of applications of classification

14
00:00:41,800 --> 00:00:46,400
from medicine to marketing to business
and lots of different areas.

15
00:00:47,100 --> 00:00:48,266
Let's have a look at a few.

16
00:00:48,266 --> 00:00:51,266
So for example,
you have customers of a business

17
00:00:51,466 --> 00:00:54,600
and you would like to predict
which ones are likely to stay

18
00:00:54,800 --> 00:00:56,633
and which ones are likely to leave.

19
00:00:56,633 --> 00:00:58,966
This is also called churn modeling.

20
00:00:58,966 --> 00:01:00,600
Very important

21
00:01:00,600 --> 00:01:03,933
because if you can predict
which customers are likely to leave

22
00:01:03,933 --> 00:01:08,333
your business in the next month
or six months, then you can take actions

23
00:01:08,533 --> 00:01:12,733
and send them special offers,
or ask them about their feedback

24
00:01:12,900 --> 00:01:15,433
and make certain changes
so that they stay.

25
00:01:15,433 --> 00:01:18,433
So it's, 
very powerful tool for businesses.

26
00:01:18,700 --> 00:01:21,366
Another application is, email.

27
00:01:21,366 --> 00:01:25,733
So for instance, if you get an email,
it might be classified

28
00:01:25,733 --> 00:01:29,466
as normal mail or it might be classified
as important an urgent mail.

29
00:01:29,466 --> 00:01:31,866
It might have a special marker,
especially if using Gmail.

30
00:01:31,866 --> 00:01:34,866
You'll see those little, chevrons

31
00:01:34,866 --> 00:01:38,300
or triangles at the beginning of a message
saying that it's important.

32
00:01:38,666 --> 00:01:42,066
Or it might be categorized as a promotion
and it might be put into a

33
00:01:42,066 --> 00:01:43,033
separate folder.

34
00:01:43,033 --> 00:01:47,066
So it's not a filling up your main, inbox
which you use for work.

35
00:01:47,400 --> 00:01:49,866
Or it might be categorized as spam.

36
00:01:49,866 --> 00:01:52,700
So spam filters have become very good.

37
00:01:52,700 --> 00:01:56,700
over the past few years,
we barely see any spam in our inboxes.

38
00:01:57,133 --> 00:02:00,566
and that is thanks to classification
algorithms.

39
00:02:01,400 --> 00:02:04,700
And another,
application is image recognition.

40
00:02:04,800 --> 00:02:08,000
so, for instance, here
we have images of dogs and cats.

41
00:02:08,266 --> 00:02:12,800
classification
would be able to separate dogs, from cats.

42
00:02:13,133 --> 00:02:15,266
So those are just a few examples.

43
00:02:15,266 --> 00:02:16,966
As you can see, it's a very powerful tool.

44
00:02:16,966 --> 00:02:20,100
There are lots of algorithms that fall
into this family of classification

45
00:02:20,100 --> 00:02:24,033
in machine learning,
and they have a variety of applications.

46
00:02:24,266 --> 00:02:27,066
So I'm very excited to explore
this section further with you together.

47
00:02:27,066 --> 00:02:29,466
And until next
time, enjoy machine learning.
