Original subtitles

1

Hello and welcome back to the course on artificial intelligence.

2

In today's section we're tackling the topic of a deep Q learning.

3

So let's see how we're going to attack this in this section.

4

We will learn a deep learning intuition the learning side of things so we go into separate deep learning

5

the intuition behind it into two parts to learning and acting.

6

And we're going to have to details on that.

7

So first of all understand how the neural networks actually learn and how they update their weights

8

based on what we are feeding them in and how the whole concept of learning works or how we're going

9

to take the temporal difference concepts or rediscuss and simple learning and apply them into deep learning

10

and then we're going to talk about how deep learning algorithm actually decide what action to take in

11

what states we're going to talk about experience replay a very important addition on top of deep learning

12

which actually enables deep you'll learn to work properly and you'll see why it's important from that

13

tutorial.

14

And then we're going to talk about action selection policies.

15

We're going to talk about how deep kill earning agents are able to combine exploration with exploitation.

16

So once they found something a good approach they can use that approach but also they need to explore

17

so that they don't get stuck in a local maximum.

18

And one more thing I wanted to mention about the section is it is highly beneficial if you have a look

19

at an x number one artificial neural networks.

20

So if you go and explore all those topics we've got some very powerful intuition tutorials prepared

21

for you there.

22

If you haven't done of course if you haven't done the deep learning course if you've done that you plan

23

of course and you already know all of these things and you can proceed with S..

24

But if you want to get that additional knowledge about neural networks before you proceed with this

25

part of the course this is highly advisable because it will help you understand exactly how neural networks

26

work and why they're so powerful why we're leveraging them in this deep cool learning algorithm.

27

And once you refresh your knowledge or gain that knowledge on neural networks from that annex and come

28

back here and we will proceed with particularly if you're pretty comfortable in neural networks then

29

let's get straight into it.

30

Let's start talking about deep learning intuition.

31

And I look forward to seeing on the first tutorial.

32

Until then enjoy a.