In this chapter, you will learn about the main concepts behind a hypercube-based NEAT algorithm and about the main challenges it was designed to solve. We take a look at the problems that arise when attempting to use direct genome encoding with large-scale artificial neural networks (ANN) and how they can be solved with the introduction of an indirect genome encoding scheme. You will learn how a Compositional Pattern Producing Network (CPPN) can be used to store genome encoding information with an extra-high compression rate and how CPPNs are employed by the HyperNEAT algorithm. Finally, you will work with practical examples that demonstrate the power of the HyperNEAT algorithm.
In this chapter, we discuss the following topics:
- The problem with the direct encoding of large-scale natural networks using NEAT, and how HyperNEAT can...
