November 2018
Intermediate to advanced
322 pages
7h 54m
English
The following diagram illustrates the structure of the encoder used for modeling. Note that it shows the MNIST digit image as an input, but we are using the Fashion-MNIST data as an input to the model:

The encoder essentially takes an input of a 28x28 image and produces a 16-dimensional representation of that image. As mentioned previously, the length of the 16D vector denotes the probability that an object is present in the image. The components of the vector represent various instantiation parameters.
The three layers dedicated to the encoder are as follows:
Read now
Unlock full access