1
00:00:00,000 --> 00:00:04,684
[MUSIC]

2
00:00:04,684 --> 00:00:06,040
Wow, exciting.

3
00:00:06,040 --> 00:00:07,890
We got through our first module.

4
00:00:07,890 --> 00:00:09,920
We got through a representation for

5
00:00:09,920 --> 00:00:14,450
linear classifier, in much more depth
than we did in the first course.

6
00:00:14,450 --> 00:00:17,800
In fact, we had a chance to dive in and

7
00:00:17,800 --> 00:00:21,630
really understand how these models are
built, the parameters and then the length

8
00:00:21,630 --> 00:00:25,640
foundation associated with them, as well
as how to extend it to other settings.

9
00:00:25,640 --> 00:00:31,480
So now we're able to take
our learning loop that goes

10
00:00:31,480 --> 00:00:37,120
from training data to creating features
h(x) to this logistic regression model.

11
00:00:37,120 --> 00:00:42,360
And define a learning algorithm which
we've talked in high level only now but

12
00:00:42,360 --> 00:00:45,980
we're going to talk in more detail in
the next module, to come up with w hat and

13
00:00:45,980 --> 00:00:47,330
we're going to make predictions from it.

14
00:00:48,700 --> 00:00:53,790
So, now you can take everything you've
learned and be able to talk about

15
00:00:53,790 --> 00:00:58,460
what does a linear classifier look like,
what's the logistic reaction model,

16
00:00:58,460 --> 00:01:03,080
how that relates to the general score
function, like we have with regression,

17
00:01:03,080 --> 00:01:08,290
but squeezes it to the interval 0,
1 to predict probabilities.

18
00:01:08,290 --> 00:01:12,750
How it's affected by parameters, and how
to extend that to multi class setting and

19
00:01:12,750 --> 00:01:14,800
how you do with category called data.

20
00:01:14,800 --> 00:01:16,390
So, we done a lot.

21
00:01:16,390 --> 00:01:21,993
We're really ready to really
think deeply about classifiers.

22
00:01:21,993 --> 00:01:22,493
[MUSIC]