﻿1
00:00:00,300 --> 00:00:01,300
Hello and welcome back.

2
00:00:01,300 --> 00:00:04,300
Let's have a look at simple linear
regression.

3
00:00:04,533 --> 00:00:05,833
So here's the equation.

4
00:00:05,833 --> 00:00:10,100
And we will look at the parts
of this equation one by one.

5
00:00:10,100 --> 00:00:14,000
So on the left we have our dependent
variable which we're trying to predict.

6
00:00:14,366 --> 00:00:17,966
On the right we have our independent
variable which is the predictor.

7
00:00:18,500 --> 00:00:23,233
here we have b0 which is the y intercept
also known as a constant.

8
00:00:23,566 --> 00:00:26,266
And b1 is the slope coefficient.

9
00:00:26,266 --> 00:00:30,233
Now to make things more fun
we're going to use that example

10
00:00:30,233 --> 00:00:33,900
we mentioned about predicting
the output of potatoes

11
00:00:33,900 --> 00:00:37,033
on a farm based on the amount
of fertilizer that we use.

12
00:00:37,500 --> 00:00:38,766
So here's our equation.

13
00:00:38,766 --> 00:00:42,433
And if we modify it to fit our example
it will look like this.

14
00:00:43,000 --> 00:00:46,433
And let's say that we ran
the simple linear regression algorithm.

15
00:00:46,433 --> 00:00:50,866
And it came up with the following values
B0 equals eight tons

16
00:00:50,866 --> 00:00:54,000
and b1 equals three tons per kilogram.

17
00:00:54,900 --> 00:00:58,900
So what does this mean
in terms of the graphical representation.

18
00:00:58,900 --> 00:01:02,100
How to how can we better understand
this on an intuitive level.

19
00:01:02,433 --> 00:01:06,200
So let's plot a simple scatter plot.

20
00:01:06,200 --> 00:01:11,133
So here we have on the x axis
a nitrogen fertilizer used in kilograms

21
00:01:11,133 --> 00:01:14,533
as our x1, variable.

22
00:01:14,533 --> 00:01:18,500
And here we have the Y variable,
which is a potato yield in tons.

23
00:01:18,933 --> 00:01:23,066
And here on the scatter plot
we have several data points.

24
00:01:23,266 --> 00:01:24,266
What are these data points.

25
00:01:24,266 --> 00:01:28,533
Well each one represents
a separate harvest on the farm

26
00:01:28,533 --> 00:01:29,666
that we are talking about.

27
00:01:29,666 --> 00:01:34,666
So, multiple times, 
the potatoes were harvested over

28
00:01:34,666 --> 00:01:39,933
many years, and the farmer recorded
how much fertilizer they used

29
00:01:39,933 --> 00:01:45,400
and also how many potatoes they were able
to harvest in that specific season.

30
00:01:45,933 --> 00:01:48,000
So there is a scatter plot.

31
00:01:48,000 --> 00:01:51,600
And what this, equation on the left
represents

32
00:01:51,600 --> 00:01:54,966
is a sloped line
that goes through a point.

33
00:01:55,400 --> 00:01:57,500
the y intercept is over here.

34
00:01:57,500 --> 00:01:59,100
That's our eight tons.

35
00:01:59,100 --> 00:02:02,800
And what the slope coefficient means
is that if you increase

36
00:02:02,800 --> 00:02:05,900
the amount of nitrogen fertilizer
by one kilogram,

37
00:02:06,366 --> 00:02:09,100
then the amount of potato output

38
00:02:09,100 --> 00:02:12,100
will increase by three tons.

39
00:02:12,466 --> 00:02:16,300
And of course, these numbers
are made up for illustrative purposes.

40
00:02:16,300 --> 00:02:18,833
So there we go. That's
how the simple linear regression works.

41
00:02:18,833 --> 00:02:20,133
I look forward to seeing you next time.

42
00:02:20,133 --> 00:02:22,133
Until then, enjoy machine learning.
