Hands-On Deep Learning with R

Creating target labels

Now, we create target labels for all the images using the following steps:

1. We need to create a matrix of labels; this is simply a matrix of binary values. We add 20 rows with values of 1 to label the fake images and 20 rows with values of 0 to label the real images. We create our matrix of label data using the following code:

labels <- rbind(matrix(1, nrow = 20, ncol = 1),                matrix(0, nrow = 20, ncol = 1))

2. After running this code, let's click on the labels object to view it. We can see that it does contain 20 rows with a value of 1 and 20 rows with a value of 0. The following is an image that you will see near the midpoint when viewing this matrix:

3. The next step is to add some noise to the labels. Just like using ...