Hands-On Neuroevolution with Python

Visual discriminator test environment

First we need to define the test environment and provide access to the dataset, which contains all the possible visual field configurations as described in the previous section. The dataset used in this experiment is created during the test environment initialization. We will discuss dataset creation later in this section.

The test environment has two major components:

• The data structure to maintain the visual field definition
• The test environment manager, which stores the dataset and provides a means to evaluate discriminator ANNs against it

Next, we provide a detailed description of these components.