We start our experiments using a simple maze configuration similar to the one described in the previous chapter. However, instead of the goal-oriented objective function, we use the NS optimization method to guide the neuroevolution process. We hope that with Novelty Search method it will be possible to find a successful maze solver with fewer epochs of evolution.
You can see the schema of the simple maze in the following plot:
The maze configuration is the same as in the previous chapter. However, we need to adjust the corresponding NEAT hyperparameters to meet the specifications of the NS optimization method.