1
00:00:00,000 --> 00:00:02,040
Congratulations on coming to

2
00:00:02,040 --> 00:00:03,930
the end of this second course.

3
00:00:03,930 --> 00:00:06,725
Just cause you've learned how
to build bigger confidence,

4
00:00:06,725 --> 00:00:08,740
how to implement
data augmentation,

5
00:00:08,740 --> 00:00:11,025
how to implement
transfer learning

6
00:00:11,025 --> 00:00:12,480
as well also how to use

7
00:00:12,480 --> 00:00:14,850
multi-class cross size but

8
00:00:14,850 --> 00:00:16,620
there's still a lot
more to learn.

9
00:00:16,620 --> 00:00:18,390
Lot's the examples you've seen so

10
00:00:18,390 --> 00:00:20,805
far have used
many computer vision.

11
00:00:20,805 --> 00:00:23,205
In the next course, you'll
learn how to deal with

12
00:00:23,205 --> 00:00:26,505
natural language processing
to how to work with texts.

13
00:00:26,505 --> 00:00:27,960
It's going to be a lot
of fun switching

14
00:00:27,960 --> 00:00:29,460
gears as well from dealing with

15
00:00:29,460 --> 00:00:32,805
pixels to dealing with characters
and dealing with words.

16
00:00:32,805 --> 00:00:35,360
We'll take a look at
how to tokenize words,

17
00:00:35,360 --> 00:00:36,830
how to generate embeddings,

18
00:00:36,830 --> 00:00:38,450
so that we can learn
off of embeddings.

19
00:00:38,450 --> 00:00:40,310
An embedding is where we can turn

20
00:00:40,310 --> 00:00:42,170
a word into basically a vector in

21
00:00:42,170 --> 00:00:43,970
a multi-dimensional space and

22
00:00:43,970 --> 00:00:45,140
from the direction that vector

23
00:00:45,140 --> 00:00:46,850
points in we can
start ascertaining

24
00:00:46,850 --> 00:00:48,455
the semantics of that word.

25
00:00:48,455 --> 00:00:50,390
We'll be going into
all of that and how

26
00:00:50,390 --> 00:00:52,610
words also work in sequence and

27
00:00:52,610 --> 00:00:55,730
different sequence
models for learning what

28
00:00:55,730 --> 00:00:57,260
the context of
a sentences and what

29
00:00:57,260 --> 00:00:59,100
the semantics of
the sentence is, yeah.

30
00:00:59,100 --> 00:01:00,755
That sounds exciting. So I think

31
00:01:00,755 --> 00:01:02,630
natural language
processing is really

32
00:01:02,630 --> 00:01:04,925
taking off partly because
of deep learning.

33
00:01:04,925 --> 00:01:06,470
So in the next course,

34
00:01:06,470 --> 00:01:08,360
you'll learn a lot about that and

35
00:01:08,360 --> 00:01:10,505
give that built some of
these exciting models yourself.

36
00:01:10,505 --> 00:01:12,930
So let's go to the next course.