1
00:00:00,810 --> 00:00:08,920
Why not pie throughout the section we're going to explore this popular library called mom pie which

2
00:00:08,920 --> 00:00:11,840
are very similar to Python lists.

3
00:00:11,860 --> 00:00:17,440
It's one of the most used and popular libraries when it comes to data science and machine learning.

4
00:00:17,440 --> 00:00:22,960
But I want to ask you why do we need this don't we have Python lists.

5
00:00:23,110 --> 00:00:29,040
Why is not pie so popular and a must use tool for machine learning.

6
00:00:29,050 --> 00:00:34,660
There are two main reasons why we want to use num PI for our problems.

7
00:00:34,750 --> 00:00:42,610
You see under the hood many functionalities that num pi offers us are actually written in C a programming

8
00:00:42,610 --> 00:00:45,480
language that is really really fast.

9
00:00:45,610 --> 00:00:51,040
If we used Python lists it would actually be a lot slower to do some of the things that we're about

10
00:00:51,040 --> 00:00:59,260
to do so num pi operations on arrays and lists are a lot faster than performing the same operations

11
00:00:59,500 --> 00:01:03,540
in Python and as you'll find out when it comes to machine learning.

12
00:01:03,540 --> 00:01:05,520
We do a lot of computation.

13
00:01:05,730 --> 00:01:12,810
We do a lot of processing and that's why GP use or graphical processing units are used because so many

14
00:01:12,810 --> 00:01:14,420
calculations need to happen.

15
00:01:14,490 --> 00:01:21,990
So we need to have any possible performance advantage we can which num pi offers the second reason we

16
00:01:21,990 --> 00:01:26,770
want to use something like num pi is that when it comes to machine learning.

17
00:01:27,240 --> 00:01:34,290
Well machines usually only understand ones and zeros and things like numbers which can be contained

18
00:01:34,290 --> 00:01:41,810
and num pi are really easy for machines to understand versus something more complicated.

19
00:01:41,860 --> 00:01:49,870
For example an image can be represented in numbers where a number represents the colour of a pixel.

20
00:01:49,870 --> 00:01:56,590
And we'll see later on the course where we try and convert different fields different attributes into

21
00:01:56,590 --> 00:01:57,530
numbers.

22
00:01:57,580 --> 00:02:06,720
For example if a patient has heart disease instead of yes or no we convert it to 1 or 0 this idea of

23
00:02:06,720 --> 00:02:13,770
converting data into something that a machine can understand like numbers like binary like ones in 0

24
00:02:14,220 --> 00:02:22,440
allows us to make machines learn that's how machines learn machines learn by performing multiple operations

25
00:02:22,890 --> 00:02:30,900
and by storing this information in an array like num pi is a great way for us to represent that data

26
00:02:31,080 --> 00:02:37,380
from what humans can understand to what computers can understand but you'll see that as we progressed

27
00:02:37,380 --> 00:02:40,170
more and more through the course for now let's get started.