WEBVTT

00:00.080 --> 00:00.880
Welcome back.

00:01.080 --> 00:06.040
Now let's print a summary of this model here.

00:06.080 --> 00:11.160
Under the compilation print model summary model.

00:11.320 --> 00:12.360
Plot summary.

00:12.480 --> 00:13.200
Run.

00:13.240 --> 00:17.920
If you have any error, go to runtime and click run all.

00:17.960 --> 00:21.480
Maybe sometimes it will be disconnected.

00:21.720 --> 00:28.600
Or maybe you restart your computer or join the class later on and you need to continue with that notebook.

00:28.720 --> 00:31.720
So notebook will maybe disconnect.

00:31.760 --> 00:37.840
Okay, so I prefer to use run all and then run this cell.

00:37.880 --> 00:40.280
Here we have the model summary.

00:40.280 --> 00:49.240
The model sequential which refers for a very simple linear regression model the linear type which is

00:49.280 --> 00:50.040
dense.

00:50.280 --> 00:53.520
The name you gave the layer dense.

00:53.520 --> 00:57.000
The type of the layer which is fully connected.

00:57.000 --> 00:59.760
And this is your only layer.

00:59.800 --> 01:01.280
The entire model.

01:01.440 --> 01:03.320
The output shape.

01:03.480 --> 01:05.280
None and one.

01:05.320 --> 01:06.080
None.

01:06.120 --> 01:10.640
Batch size can vary any number of samples one.

01:10.680 --> 01:14.520
Each prediction is a single number the price.

01:14.680 --> 01:24.280
So one house one 110 houses ten one okay, because each prediction is a single number, which is the

01:24.280 --> 01:32.410
price if we have 1010 houses, we need to get the only one thing to be predicted, which is the price.

01:32.450 --> 01:36.050
Okay, param not total parameters.

01:36.290 --> 01:36.810
Do.

01:36.970 --> 01:39.570
This is the most important part.

01:39.610 --> 01:42.650
What are those two parameters?

01:42.850 --> 01:46.570
The slope and the intercept.

01:46.610 --> 01:48.090
W and b.

01:48.330 --> 02:00.130
So here if we go to our model here we have the price equals to 2.9 times size plus 2.7.

02:00.490 --> 02:03.010
The slope is the size.

02:03.250 --> 02:06.330
The intercept is b here okay.

02:06.410 --> 02:09.530
This is a general formula for our model.

02:09.530 --> 02:16.050
This is not exactly this because price equals to 2.9 times size plus 2.7.

02:16.090 --> 02:19.730
This is from the previous example and previous model.

02:19.730 --> 02:30.330
But in our application we have exactly similar to what we've learned y equal to ax plus b a the slope

02:30.330 --> 02:32.930
and b the intercept okay.

02:33.130 --> 02:41.410
So those are the two parameters that our machine learning model we're going to learn.

02:41.450 --> 02:44.330
For that we have two parameters okay.

02:44.570 --> 02:47.570
This is how we get a model summary.