ABSTRACT
Genetic algorithms and genetic programming are genetics-based optimization methods in which potential solutions evolve via operators such as selection, crossover, and mutation. Logic-based neural networks are a variation of standard neural networks; they fill the gap between distributed, unstructured neural networks and symbolic programming. The algorithm is meant to be part of a two-level development process where, at first, satisfactory logic-based neural networks are obtained using our algorithm; then, gradient-based learning methods are used to refine the networks. The chapter introduces modifications to the genetic programming paradigm (GPP) that make it possible for the GPP to obtain Logic-based neural networks (LNN). The modifications include a new, more general data structure which is very suitable to LNNs; also, new operators that are suitable to LNNs are developed. The LNNs obtained using the GPP method are starting architectures from which more refined, gradient-based learning is performed using LNN-adapted learning methods.
