1
00:00:00,450 --> 00:00:04,030
Now we've got a framework we can use for machine learning problems.

2
00:00:04,080 --> 00:00:10,450
It's time to start matching the framework to existing data science and machine learning tools.

3
00:00:10,470 --> 00:00:15,960
We're going to start right at the foundation says machine learning is quite broad.

4
00:00:15,960 --> 00:00:18,550
There are many different tools that we could use.

5
00:00:18,600 --> 00:00:25,050
So we need something to help manage our tools for us to do so.

6
00:00:25,200 --> 00:00:32,380
You can use Anaconda many condo and condo and now I know you might be thinking snakes.

7
00:00:32,520 --> 00:00:34,080
What the hell is Daniel talking about.

8
00:00:34,410 --> 00:00:40,980
But trust me once you understand these setting up future machine learning projects such as ones which

9
00:00:40,980 --> 00:00:45,160
use tools like you see here are going to be a breeze.

10
00:00:46,970 --> 00:00:54,650
You can think of Anaconda and many Conda as the hardware store and workbench of data scientists and

11
00:00:54,660 --> 00:01:02,140
Conda as your personal assistant Anaconda and many conduct are software distributions.

12
00:01:02,660 --> 00:01:06,020
So much like when you download an app from the App Store.

13
00:01:06,140 --> 00:01:07,250
It comes with code.

14
00:01:07,250 --> 00:01:12,350
Other people have written which you can use on your computer or smartphone.

15
00:01:12,350 --> 00:01:14,180
The same goes for these two.

16
00:01:14,240 --> 00:01:15,470
They come with code.

17
00:01:15,470 --> 00:01:21,830
Other people have written which we can take advantage of on our computers for our own machine learning

18
00:01:21,830 --> 00:01:29,720
projects collections of code written by other people are often referred to as packages or tools.

19
00:01:29,720 --> 00:01:35,600
For example for our heart disease classification problem we've talked about we might want to use several

20
00:01:35,600 --> 00:01:42,480
Python packages each of which could be considered a different tool for different stages of the project.

21
00:01:42,620 --> 00:01:50,330
In the beginning we might use pandas and map plot leave packages for data analysis and data exploration

22
00:01:50,930 --> 00:01:57,500
then to build a machine learning model we might use the psychic loan package it might use.

23
00:01:57,530 --> 00:02:06,670
All of these packages within a Jupiter notebook which comes from the Jupiter package now downloading

24
00:02:06,670 --> 00:02:13,930
Anaconda into your computer requires around three gigabytes or so of space and it installs all of the

25
00:02:13,930 --> 00:02:17,750
major data science and machine learning packages.

26
00:02:17,800 --> 00:02:24,070
However remember how I said Anaconda is like the hardware store of data science tools.

27
00:02:24,070 --> 00:02:27,790
This means it's got almost every data science tool you can imagine.

28
00:02:28,540 --> 00:02:32,000
And from my experience many of them don't get used.

29
00:02:32,350 --> 00:02:38,890
And since we're focused on being practitioners we're going to be using the most useful ones.

30
00:02:38,910 --> 00:02:40,990
This is where many contact comes in.

31
00:02:41,290 --> 00:02:47,800
Many come to remember is that the data scientists workbench it starts with the minimum requirements

32
00:02:47,920 --> 00:02:51,390
to get started such as Python and a few others.

33
00:02:51,550 --> 00:02:57,850
And now because of this it takes up up to about 10 times this space on your computer.

34
00:02:57,850 --> 00:03:05,380
Coming in and around 200 megabytes so to save space and to get started we're going to use many Conda.

35
00:03:05,740 --> 00:03:09,430
But I mentioned Anaconda because I want you to be aware of it.

36
00:03:09,460 --> 00:03:13,460
If someone mentions it or you see it somewhere online.

37
00:03:13,560 --> 00:03:19,080
Up next we're going to look at Conda which is like your assistant that comes with many conduct.

38
00:03:19,230 --> 00:03:25,050
So once you download many conduit we'll have a few tools but then you use Conda to help manage those

39
00:03:25,050 --> 00:03:26,890
tools for you.

40
00:03:26,950 --> 00:03:28,230
We'll look into this more in a second.