Model building

We're finally ready to start building our CNN in Keras. In this section, we'll take two different approaches to model building. First, we'll start by building a relatively simple CNN consisting of a few layers. We'll take a look at the performance of the simple model, and discuss its pros and cons. Next, we'll use a model that was considered state-of-the art just a few years ago—the VGG16 model. We'll see how we can leverage on the pre-trained weights to adapt the VGG16 model for cats versus dogs image classification.

Building a simple CNN

In an earlier section, we showed how the fundamental building blocks of a CNN consist of a series of convolutional and pooling layers...