1
00:00:00,319 --> 00:00:02,190
Congratulations on completing this

2
00:00:02,190 --> 00:00:04,830
course! Let's quickly recap what we

3
00:00:04,830 --> 00:00:08,730
covered in this course. In module 1, we

4
00:00:08,730 --> 00:00:10,559
went over a number of exciting and

5
00:00:10,559 --> 00:00:12,500
motivating applications of deep learning.

6
00:00:12,500 --> 00:00:15,269
We also briefly covered neurons and

7
00:00:15,269 --> 00:00:17,100
neural networks in the brain to

8
00:00:17,100 --> 00:00:19,410
appreciate how they inspire artificial

9
00:00:19,410 --> 00:00:21,869
neural networks, and then we learned how

10
00:00:21,869 --> 00:00:24,119
neural networks make predictions through

11
00:00:24,119 --> 00:00:26,330
the forward propagation process. In

12
00:00:26,330 --> 00:00:29,550
module 2, we learned how an artificial

13
00:00:29,550 --> 00:00:31,740
neural network learns through gradient

14
00:00:31,740 --> 00:00:34,170
descent and backpropagation. We also

15
00:00:34,170 --> 00:00:35,670
learned about the vanishing gradient

16
00:00:35,670 --> 00:00:38,219
problem, and which activation functions

17
00:00:38,219 --> 00:00:41,719
are best to overcome this problem. In

18
00:00:41,719 --> 00:00:44,879
module 3, we learnt about the Keras

19
00:00:44,879 --> 00:00:47,160
library and how to use it to build

20
00:00:47,160 --> 00:00:48,800
models for regression and classification

21
00:00:48,800 --> 00:00:54,000
problems. In module 4, we learned about

22
00:00:54,000 --> 00:00:56,340
supervised and unsupervised deep neural

23
00:00:56,340 --> 00:00:58,800
networks, and we used the Keras library

24
00:00:58,800 --> 00:01:03,260
to build a convolutional neural network.

25
00:01:03,260 --> 00:01:05,080
Finally, I really

26
00:01:05,080 --> 00:01:07,260
enjoyed putting this course together, and

27
00:01:07,260 --> 00:01:09,659
so I really hope that you enjoyed it and

28
00:01:09,659 --> 00:01:11,159
learned a lot from it.

29
00:01:11,160 --> 00:01:12,820
It was difficult for me at some point,

30
00:01:12,820 --> 00:01:14,500
because I had to leave out a lot of

31
00:01:14,500 --> 00:01:17,800
details in order to keep it simple. The

32
00:01:17,800 --> 00:01:20,120
point of this course was not to teach you

33
00:01:20,120 --> 00:01:22,160
everything but to teach you enough to be

34
00:01:22,160 --> 00:01:23,900
ready for more advanced courses on

35
00:01:23,909 --> 00:01:26,100
deep learning, and to even start

36
00:01:26,100 --> 00:01:28,720
learning independently if you wish. And I

37
00:01:28,720 --> 00:01:31,020
think that you are there now. Thank you

38
00:01:31,020 --> 00:01:32,660
for taking this course and I wish you

39
00:01:32,660 --> 00:01:34,200
the best of luck!