1
00:00:00,000 --> 00:00:04,602
[MUSIC]

2
00:00:04,602 --> 00:00:05,805
What we've covered so

3
00:00:05,805 --> 00:00:11,030
far is mostly a review of what you saw in
the first course of the specialization.

4
00:00:11,030 --> 00:00:14,970
At this point,
we're going to go much deeper into

5
00:00:14,970 --> 00:00:18,270
fundamental concepts which
are associated in logistical regression.

6
00:00:18,270 --> 00:00:22,800
In particular, in logistic regression, we
don't just predict plus one or minus one.

7
00:00:22,800 --> 00:00:24,400
We predict a probability.

8
00:00:24,400 --> 00:00:27,390
How likely is this review to be positive?

9
00:00:27,390 --> 00:00:28,940
How likely is it to be negative?

10
00:00:28,940 --> 00:00:32,060
And those probabilities are extremely
useful because they give us

11
00:00:32,060 --> 00:00:36,220
an indication of how sure we
are about predictions we make.

12
00:00:36,220 --> 00:00:37,700
So, in particular, so

13
00:00:37,700 --> 00:00:41,420
far we talked about taking the score,
throwing it into the sine function.

14
00:00:41,420 --> 00:00:47,130
Deciding is it a positive or
a negative review, so plus 1 or minus1.

15
00:00:47,130 --> 00:00:50,750
However, not all reviews
are created the same.

16
00:00:50,750 --> 00:00:54,050
So, for example, Rita says,
the sushi & everything was awesome!

17
00:00:55,070 --> 00:00:58,881
That's a definite plus 1,
I feel really good about that.

18
00:00:58,881 --> 00:01:00,750
But if Rita ays, the sushi was good,

19
00:01:00,750 --> 00:01:06,760
the service was okay,
I don't know what people think about that.

20
00:01:06,760 --> 00:01:08,200
It could be plus 1, it could be minus 1.

21
00:01:08,200 --> 00:01:11,820
If the person is really nice they might
feel bad about writing negative reviews,

22
00:01:11,820 --> 00:01:14,980
they might write that when they
really mean it's a negative review.

23
00:01:14,980 --> 00:01:19,250
So how can I capture the fact that for
the first one I'm sure

24
00:01:19,250 --> 00:01:23,200
that Y hat is plus one while for
the one on the right I might say.

25
00:01:25,450 --> 00:01:29,984
Y hat I don't know if its plus one or
minus one so I'm going to say Y hat

26
00:01:29,984 --> 00:01:34,047
is plus one probability 0.5 so
its either way its fine.

27
00:01:34,047 --> 00:01:37,619
[MUSIC]