1
00:00:00,463 --> 00:00:03,933
[MUSIC]

2
00:00:03,933 --> 00:00:06,268
I've been doing machine learning for
a long time, and

3
00:00:06,268 --> 00:00:09,061
it's a really exciting time to
be doing machine learning today,

4
00:00:09,061 --> 00:00:12,020
because we're really seeing
a lot of impact from it.

5
00:00:12,020 --> 00:00:14,340
But this is not how I
started machine learning.

6
00:00:14,340 --> 00:00:16,440
If you just talked to me a few years ago,

7
00:00:16,440 --> 00:00:19,380
the way I thought about machine
learning was quite different.

8
00:00:19,380 --> 00:00:22,340
You started with some data sets,
some data.

9
00:00:22,340 --> 00:00:26,920
And you fed it to some magical
machine learning algorithm and

10
00:00:26,920 --> 00:00:32,110
then I just showed you that my curve
was better than your curve and

11
00:00:32,110 --> 00:00:34,090
then I wrote the paper to
machine learning conference.

12
00:00:35,590 --> 00:00:37,900
And that's how I always thought
about machine learning.

13
00:00:37,900 --> 00:00:40,110
But this is not why I got
into the first place,

14
00:00:40,110 --> 00:00:41,970
why did I get into machine learning?

15
00:00:41,970 --> 00:00:44,320
Now when I was a kid I
read a lot of books.

16
00:00:44,320 --> 00:00:45,110
A lot of sci-fi.

17
00:00:45,110 --> 00:00:50,800
You know I really wanted to build robots
that were really intelligent that

18
00:00:50,800 --> 00:00:53,640
really thought about the world and

19
00:00:53,640 --> 00:00:57,900
reasoned about things, what I call
today intelligent applications.

20
00:00:57,900 --> 00:01:02,490
And the really cool and exciting thing
is that today we're seeing a lot of

21
00:01:02,490 --> 00:01:07,020
impact from intelligent applications,
they're using machine learning.

22
00:01:07,020 --> 00:01:12,440
In fact if you look at English
industry successful companies today.

23
00:01:12,440 --> 00:01:16,720
Companies that are called disruptive, they
take a market and completely change it.

24
00:01:16,720 --> 00:01:21,153
They're often differentiated
by intelligent applications,

25
00:01:21,153 --> 00:01:25,630
by intelligence that uses
machine learning at its core.

26
00:01:25,630 --> 00:01:29,550
So, for example, early days Amazon
really disrupted the retail market

27
00:01:29,550 --> 00:01:32,670
by bringing in product
recommendations into their website.

28
00:01:32,670 --> 00:01:38,790
We saw Google disrupting the advertising
market by really targeting advertising

29
00:01:38,790 --> 00:01:44,020
with machine learning to figure
out what people would click on.

30
00:01:44,020 --> 00:01:48,300
You saw Netflix,
the movie distribution company,

31
00:01:48,300 --> 00:01:51,800
really change how movies are seen.

32
00:01:51,800 --> 00:01:56,000
Now we don't go to a shop and
rent movies anymore.

33
00:01:56,000 --> 00:01:58,210
We go to the web and we stream data.

34
00:01:58,210 --> 00:01:59,380
Netflix really changed that.

35
00:01:59,380 --> 00:02:03,390
And at the core there was a recommender
system that helped me find the movies

36
00:02:03,390 --> 00:02:06,970
that I liked, the movies that are good for
me out of the many, many,

37
00:02:06,970 --> 00:02:10,480
many thousands of movies
they were serving.

38
00:02:10,480 --> 00:02:14,470
You see companies like Pandora,
where they're providing

39
00:02:14,470 --> 00:02:19,340
a music recommendation system
where I find music that I like.

40
00:02:19,340 --> 00:02:23,320
And I find streams that are good for
the morning when I'm sleepy or

41
00:02:23,320 --> 00:02:27,540
at night when I'm ready to go to bed and
I want to listen to different music.

42
00:02:27,540 --> 00:02:29,960
And they really find good music for us.

43
00:02:29,960 --> 00:02:32,860
And you see that in many places,
in many industries,

44
00:02:32,860 --> 00:02:37,380
you see Facebook connecting me with people
who I might want to be friends with.

45
00:02:37,380 --> 00:02:42,280
And you even see companies like Uber
disrupting the taxi industry by really

46
00:02:42,280 --> 00:02:46,530
optimizing how to connect drivers
with people in real time.

47
00:02:46,530 --> 00:02:51,480
So, in all these areas, machine learning
is one of the core technologies,

48
00:02:51,480 --> 00:02:55,920
the technology that makes that
company's product really special.

49
00:02:55,920 --> 00:03:00,150
So in this specialization, we're really
going to talk about all aspects of machine

50
00:03:00,150 --> 00:03:04,528
learning and get you ready to build
intelligent applications just like that.

51
00:03:04,528 --> 00:03:07,770
And we're gonna see a pipeline again,
and again, and

52
00:03:07,770 --> 00:03:12,420
again where we're gonna start from data
and really bring in a machine learning

53
00:03:12,420 --> 00:03:16,370
method that provides you with a new
kind of analysis of the data.

54
00:03:16,370 --> 00:03:18,840
And that analysis is gonna
give you intelligence.

55
00:03:18,840 --> 00:03:23,580
Intelligence like what product
am I likely to buy right now?

56
00:03:23,580 --> 00:03:27,568
And by taking this pipeline and working
through it in a wide range of settings and

57
00:03:27,568 --> 00:03:31,855
a wide range of applications, with a range
of algorithms, but really understanding

58
00:03:31,855 --> 00:03:35,547
how they connect together, you're
gonna be able to really build smart,

59
00:03:35,547 --> 00:03:37,595
intelligent applications of your own.

60
00:03:37,595 --> 00:03:38,095
[MUSIC]