1
00:00:00,000 --> 00:00:04,338
[MUSIC]

2
00:00:04,338 --> 00:00:08,690
Okay, so in this module we've talked
about simple linear regression.

3
00:00:08,690 --> 00:00:12,100
And what we've seen is we've
described what the model is.

4
00:00:12,100 --> 00:00:16,620
We just have a single input, single
output, and fitting a line or model.

5
00:00:16,620 --> 00:00:21,810
It's just a simple line, to describe
the relationship between our input x and

6
00:00:21,810 --> 00:00:23,320
our output y.

7
00:00:23,320 --> 00:00:28,780
We've talked about goodness of fit of a
specific line to our data and the measure

8
00:00:28,780 --> 00:00:32,450
being the residual sum of squares that
we've talked about in this module.

9
00:00:32,450 --> 00:00:36,360
And we've also talked about
some ways to think about

10
00:00:36,360 --> 00:00:40,420
interpreting our fitted line and
using it to form predictions.

11
00:00:40,420 --> 00:00:44,820
But a big emphasis was on thinking
about how do we actually fit

12
00:00:44,820 --> 00:00:50,070
that line to the data, and we talked
about different optimization techniques.

13
00:00:50,070 --> 00:00:55,630
The big one being, gradient descent,
and using that to minimize our residual

14
00:00:55,630 --> 00:01:00,720
sum squares, to come up with our fitted
line that we're gonna use for predictions.

15
00:01:00,720 --> 00:01:03,450
Even though this is a very very simple and

16
00:01:03,450 --> 00:01:07,350
basic tool,
it's actually incredibly powerful.

17
00:01:07,350 --> 00:01:11,046
And we'll look at this in
some of our assignments.

18
00:01:11,046 --> 00:01:16,369
[MUSIC]