Neuro-Symbolic Programming in Python

Neuro-symbolic artificial intelligence (NSAI) systems are not constrained by standardized principles or confined by special requirements. The only consideration in NSAI systems is the combination of symbolic artificial intelligence (AI) and neural networks (NNs). As reviewed in the previous chapter, NSAI implementations vary widely in scope and architecture. Although NSAI is a product of decades of research, the marriage between the two fields remains fresh. Some frameworks have gained more traction than others; however, none stand out as state of the art yet. Even evaluation strategies focused on NSAI algorithms are yet to solidify. Designing NSAI solutions requires solid creative direction by the developers, heavily influenced by domain knowledge and expertise. When working with NSAI, there are some considerations to keep in mind. How will deep learning interact (either directly or indirectly) with the symbolic learning component? How will we...