0
1
00:00:01,500 --> 00:00:10,080
We have all hearing these Buzz words recently, like artificial intelligence, machine learning, A.I. , Deep learning.
1

2
00:00:10,710 --> 00:00:17,640
but there is a lack of clear distinction due to the tightly intertwined terminologies.
2

3
00:00:17,850 --> 00:00:24,630
Let me share with simple definition, explanation and sharp differentiation between machine learning,
3

4
00:00:25,020 --> 00:00:29,140
deep learning and artificial intelligence terminologies.
4

5
00:00:29,280 --> 00:00:35,520
Here we can see we have this artificial intelligence, which is a technique that enables computers to
5

6
00:00:35,520 --> 00:00:41,320
mimic human behaviors and the subset of these artificial intelligence.
6

7
00:00:41,550 --> 00:00:44,340
We have machine learning. 
7

8
00:00:44,340 --> 00:00:52,710
We can define it as a subset of A.I. techniques which use statistical methods to enable machines to
8

9
00:00:52,710 --> 00:00:55,170
improve with experience.
9

10
00:00:55,320 --> 00:01:00,210
Under this machine learning, we have another subset, which is deep learning.
10

11
00:01:01,020 --> 00:01:10,110
Deep learning is a machine learning, which makes the computation of multilayer neural networks feasible.
11

12
00:01:10,170 --> 00:01:16,800
I hope you now understand better the difference between deep learning, machine learning and artificial
12

13
00:01:16,980 --> 00:01:17,870
intelligence.
13

14
00:01:17,880 --> 00:01:22,730
We are focusing in this course on machine learning.
14

15
00:01:22,740 --> 00:01:30,240
It's a part of the computer field that involved from pattern recognition and computation learning in artificial
15

16
00:01:30,240 --> 00:01:38,550
intelligence, which deal with building systems and algorithms that can learn from data without relying
16

17
00:01:38,550 --> 00:01:40,550
on rule based programming.
17

18
00:01:40,590 --> 00:01:42,870
Here, it's a very important point.
18

19
00:01:43,050 --> 00:01:49,020
We need to make a difference between rule based programming, which is several lines, and a program
19

20
00:01:49,020 --> 00:01:51,930
is just following certain lines and instruction.
20

21
00:01:52,050 --> 00:01:56,400
Machine learning is working with sensors and sensing the environment.
21

22
00:01:56,400 --> 00:02:02,750
Machine Learning System is able to make decision based on a new situation based on a new data.
22

23
00:02:02,940 --> 00:02:09,210
If it's an industry, then they are hearing different sensors, they're receiving input and based on
23

24
00:02:09,210 --> 00:02:11,790
the new situation, they can make a new decision.
24

25
00:02:11,880 --> 00:02:19,500
If it's only a software program with new data, they can analyze the data based on the experience.
25

26
00:02:19,500 --> 00:02:27,090
They can make a new decision. Unless the rule-based programming, which is only able to follow certain lines
26

27
00:02:27,090 --> 00:02:28,280
of instructions.
27

28
00:02:28,860 --> 00:02:33,540
There are several techniques to implement and to apply machine learning.
28

29
00:02:33,810 --> 00:02:42,750
There are linear regression, K-means decision trees, random forest, PCA, super-vector-machines,
29

30
00:02:42,750 --> 00:02:44,910
artificial neural networks (ANN).
30

31
00:02:45,240 --> 00:02:49,950
In this course, we're going to focus on artificial neural networks.